Smith v. Arizona, 602 U.S. ___ (2024)

Docket No. 22-899
Granted: September 29, 2023
Argued: January 10, 2024
Decided: June 21, 2024
Justia Summary

The case involves the Sixth Amendment's Confrontation Clause, which guarantees a criminal defendant the right to confront the witnesses against him. The defendant, Jason Smith, was charged with various drug offenses after law enforcement officers found him with a large quantity of what appeared to be drugs and drug-related items. The seized items were sent to a crime lab for scientific analysis. Analyst Elizabeth Rast ran forensic tests on the items and concluded that they contained usable quantities of methamphetamine, marijuana, and cannabis. Rast prepared a set of typed notes and a signed report about the testing. However, Rast stopped working at the lab prior to trial, so the State substituted another analyst, Greggory Longoni, to provide an independent opinion on the drug testing performed by Rast. At trial, Longoni conveyed to the jury what Rast’s records revealed about her testing, before offering his “independent opinion” of each item’s identity. Smith was convicted. On appeal, he argued that the State’s use of a substitute expert to convey the substance of Rast’s materials violated his Confrontation Clause rights. The Arizona Court of Appeals rejected Smith’s challenge.

Primary Holding

When an expert testifying at a criminal trial conveys an absent analyst’s statements in support of the expert’s opinion, and the statements provide that support only if true, the statements come into evidence for their truth.


NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U.S. 321, 337.




certiorari to the court of appeals of arizona, division one

No. 22–899. Argued January 10, 2024—Decided June 21, 2024

NOTICE: This opinion is subject to formal revision before publication in the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543,, of any typographical or other formal errors.



No. 22–899



on writ of certiorari to the court of appeals of arizona, division one

[June 21, 2024]

Justice Kagan delivered the opinion of the Court.

The Sixth Amendment’s Confrontation Clause guarantees a criminal defendant the right to confront the witnesses against him. The Clause bars the admission at trial of “testimonial statements” of an absent witness unless she is “unavailable to testify, and the defendant ha[s] had a prior opportunity” to cross-examine her. Crawford v. Washington, 541 U.S. 36, 53–54 (2004). And that prohibition applies in full to forensic evidence. So a prosecutor cannot introduce an absent laboratory analyst’s testimonial out-of-court statements to prove the results of forensic testing. See Melendez-Diaz v. Massachusetts, 557 U.S. 305, 307, 329 (2009).

The question presented here concerns the application of those principles to a case in which an expert witness restates an absent lab analyst’s factual assertions to support his own opinion testimony. This Court has held that the Confrontation Clause’s requirements apply only when the prosecution uses out-of-court statements for “the truth of the matter asserted.” Crawford, 541 U. S., at 60, n. 9. Some state courts, including the court below, have held that this condition is not met when an expert recites another analyst’s statements as the basis for his opinion. Today, we reject that view. When an expert conveys an absent analyst’s statements in support of his opinion, and the statements provide that support only if true, then the statements come into evidence for their truth. As this dispute illustrates, that will generally be the case when an expert relays an absent lab analyst’s statements as part of offering his opinion. And if those statements are testimonial too—an issue we briefly address but do not resolve as to this case—the Confrontation Clause will bar their admission.



The Confrontation Clause provides that “[i]n all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.” In operation, the Clause protects a defendant’s right of cross-examination by limiting the prosecution’s ability to introduce statements made by people not in the courtroom. For a time, this Court held that the Clause’s “preference for face-to-face” confrontation could give way if a court found that an out-of-court statement bore “adequate indicia of reliability.” Ohio v. Roberts, 448 U.S. 56, 65–66 (1980). But two decades ago, the Court changed course, to better reflect original understandings. In Crawford v. Washington, the Court deemed it “fundamentally at odds with the right of confrontation” to admit statements based on judicial determinations of reliability. 541 U. S., at 61. The Clause, Crawford explained, “commands[ ] not that evidence be reliable, but that reliability be assessed in a particular manner: by testing in the crucible of cross-examination.” Ibid. And so the Clause bars the admission at trial of an absent witness’s statements—however trustworthy a judge might think them—unless the witness is unavailable and the defendant had a prior chance to subject her to cross-examination.

But not always. The Clause’s prohibition “applies only to testimonial hearsay”—and in that two-word phrase are two limits. Davis v. Washington, 547 U.S. 813, 823 (2006). First, in speaking about “witnesses”—or “those who bear testimony”—the Clause confines itself to “testimonial statements,” a category whose contours we have variously described. Id., at 823, 826; see id., at 822 (statements “made in the course of police interrogation” were testimonial when “the primary purpose of the interrogation [was] to establish or prove past events potentially relevant to later criminal prosecution”); Michigan v. Bryant, 562 U.S. 344, 358, 359 (2011) (statements made to police “to meet an ongoing emergency” were “not procured with a primary purpose of creating an out-of-court substitute for trial testimony”); Melendez-Diaz, 557 U. S., at 311 (testimonial certificates of the results of forensic analysis were created “under circumstances which would lead an objective witness reasonably to believe that the statement[s] would be available for use at a later trial”); infra, at 19. Second and more relevant here, the Clause bars only the introduction of hearsay—meaning, out-of-court statements offered “to prove the truth of the matter asserted.” Anderson v. United States, 417 U.S. 211, 219 (1974). When a statement is admitted for a reason unrelated to its truth, we have held, the Clause’s “role in protecting the right of cross-examination” is not implicated. Tennessee v. Street, 471 U.S. 409, 414 (1985); see Anderson, 417 U. S., at 220. That is because the need to test an absent witness ebbs when her truthfulness is not at issue. See ibid.; Street, 471 U. S., at 414; infra, at 13–14, 17.

Not long after Crawford, the Court made clear that the Confrontation Clause applies to forensic reports. In Melendez-Diaz v. Massachusetts, state prosecutors introduced “certificates of analysis” (essentially, affidavits) stating that lab tests had identified a substance seized from the defendant as cocaine. 557 U. S., at 308. But the State did not call as witnesses the analysts who had conducted the tests and signed the certificates. We held that a “straightforward application” of Crawford showed a constitutional violation. 557 U. S., at 312. The certificates were testimonial: They had an “evidentiary purpose,” identical to the one served had the analysts given “live, in-court testimony.” Id., at 311. And the certificates were offered to prove the truth of what they asserted: that the seized powder was in fact cocaine. See id., at 310–311. So the defendant had a right to cross-examine the lab-analyst certifiers. In reaching that conclusion, we rejected the State’s claim that the results of so-called “neutral, scientific testing” should be subject to a different rule. Id., at 317. We again underscored that the Confrontation Clause commanded not reliability but one way of testing it—through cross-examination. See ibid. And we thought that method might have plenty to do in cases involving forensic analysis. After all, lab tests are “not uniquely immune from the risk of manipulation” or mistake. Id., at 318. The defendant might have used cross-examination to probe “what tests the analysts performed,” whether those tests “present[ed] a risk of error,” and whether the analysts had the right skill set to “interpret[ ] their results.” Id., at 320.

Two years later, the Court relied on Melendez-Diaz to hold that a State could not introduce one lab analyst’s written findings through the testimony of another. In Bullcoming v. New Mexico, 564 U.S. 647, 651–652 (2011), an analyst tested the blood-alcohol level of someone charged with drunk driving, and prepared a “testimonial certification” reporting that the level was higher than legal. But by the time the driver’s trial began, that analyst had been placed on unpaid leave. So the State instead called a different analyst from the same lab to testify as to what the certification said. The substitute analyst had similar qualifications, and knew about the type of test performed. But the Court held that insufficient to satisfy the Confrontation Clause. The “surrogate testimony,” the Court explained, “could not convey what [the certifying analyst] knew or observed” about “the particular test and testing process he employed.” Id., at 661. Nor could that “testimony expose any lapses or lies on the certifying analyst’s part,” or offer any insight into whether his leave-without-pay was the result of misconduct. Id., at 662. Concluded the Court: “[W]hen the State elected to introduce [the] certification,” its author—and not any substitute—“became [the] witness [that the defendant] had the right to confront.” Id., at 663.

The very next Term brought another case in which one lab analyst related what another had found—though this time on the way to stating her own conclusion. In Williams v. Illinois, 567 U.S. 50 (2012), state police sent vaginal swabs from a rape victim known as L. J. to a private lab for DNA testing. When the lab sent back a DNA profile, a state analyst checked it against the police department’s database and found that it matched the profile of prior arrestee Sandy Williams. The State charged Williams with the rape, and he went to trial. The prosecution chose not to bring the private lab analyst to the stand. Instead, it called Sandra Lambatos, the state analyst who had searched the police database and found the DNA match. Lambatos had no first-hand knowledge of how the private lab had produced its results; she did not even know whether those results actually came from L. J.’s vaginal swabs (as opposed to some other sample). But she spoke repeatedly about comparing Williams’s DNA to the DNA “found in [L. J.’s] vaginal swabs.” Id., at 61, 71 (plurality opinion); see id., at 124 (Kagan, J., dissenting). So in addition to describing how she discovered a match, Lambatos became the conduit for what a different analyst had reported—that a particular DNA profile came from L. J.’s vaginal swabs. Williams objected, at trial and later: He thought that, just as in Bullcoming, crucial evidence had been admitted through a surrogate expert, thus violating his right of confrontation.

But the Illinois Supreme Court rejected Williams’s claim, holding that Lambatos’s testimony about the private lab analyst’s finding did not raise a Confrontation Clause issue. See People v. Williams, 238 Ill. 2d 125, 143–144, 939 N.E.2d 268, 278–279 (2010). The court explained that under state evidence law, an expert can disclose “underlying facts and data” for “the purpose of explaining the basis for [her] opinion.” Id., at 137, 143, 939 N. E. 2d, at 274–275, 278. And when she does so, the court held, the testimony is not subject to the Confrontation Clause because it is not admitted “for the truth of the matter asserted.” Id., at 143, 939 N. E. 2d, at 278. Thus, Lambatos could relay the private lab’s finding that L. J.’s vaginal swabs produced a certain DNA profile in order to “explain[ ] the basis for her opinion” that “there was a DNA match between [Williams’s] blood sample and the semen sample recovered from L. J.” Id., at 150, 939 N. E. 2d, at 282. The admission of the private lab report’s contents for that “limited purpose,” the court reasoned, would “aid the [factfinder] in assessing the value of [Lambatos’s] opinion.” Id., at 144, 939 N. E. 2d, at 278; see id., at 150, 939 N. E. 2d, at 282.

This Court granted Williams’s petition for certiorari, but failed to produce a majority opinion. Four Members of the Court approved the Illinois Supreme Court’s approach to “basis evidence,” and agreed that Lambatos’s recitation of the private lab’s findings served “the legitimate nonhearsay purpose of illuminating the expert’s thought process.” Williams, 567 U. S., at 78 (plurality opinion). But the remaining five Members rejected that view. Those five stated, in two opinions, that basis evidence is generally introduced for its truth, and was so introduced at Williams’s trial. Justice Thomas explained that “the purportedly limited reason for [the basis] testimony—to aid the factfinder in evaluating the expert’s opinion—necessarily entail[ed] an evaluation of whether [that] testimony [was] true”: “[T]he validity of Lambatos’[s] opinion ultimately turned on the truth of [the private lab analyst’s] statements.” Id., at 106, n. 1, 108 (concurring in judgment). A dissent for another four Justices agreed: “[T]he utility of the [private analyst’s] statement that Lambatos repeated logically depended on its truth.” Id., at 132 (opinion of Kagan, J.). And the State could not avoid that conclusion by “rely[ing] on [Lambatos’s] status as an expert.” Id., at 126. Those shared views might have made for a happy majority, except that a different Confrontation Clause issue intruded. Justice Thomas thought that the private lab report was not testimonial because it lacked sufficient formality, so affirmed the Illinois Supreme Court on that alternative ground. The bottom line was that Williams lost, even though five Members of this Court rejected the state court’s “not for the truth” reasoning.[1]

Our opinions in Williams “have sown confusion in courts across the country” about the Confrontation Clause’s application to expert opinion testimony. Stuart v. Alabama, 586 U. S. ___, ___ (2018) (Gorsuch, J., dissenting from denial of certiorari) (slip op., at 2). Some courts have applied the Williams plurality’s “not for the truth” reasoning to basis testimony, while others have adopted the opposed five-Justice view.[2] This case emerged out of that muddle.


Like Melendez-Diaz, this case involves drugs. In December 2019, Arizona law enforcement officers executed a search warrant on a property in the foothills of Yuma County. Inside a shed on the property, they found petitioner Jason Smith. They also found a large quantity of what appeared to be drugs and drug-related items. As a result, Smith was charged with possessing dangerous drugs (methamphetamine) for sale; possessing marijuana for sale; possessing narcotic drugs (cannabis) for sale; and possessing drug paraphernalia. He pleaded not guilty, and the case was set for trial.

In preparation, the State sent items seized from the shed to a crime lab run by the Arizona Department of Public Safety (DPS) for a “full scientific analysis.” App. to Pet. for Cert. 127a. The State’s request identified Smith as the individual “associated” with the substances, listed the charges against him, and noted that “[t]rial ha[d] been set.” Ibid. Analyst Elizabeth Rast communicated with prosecutors about exactly which items needed to be examined, and then ran the requested tests. See id., at 99a.

Rast prepared a set of typed notes and a signed report, both on DPS letterhead, about the testing. The notes documented her lab work and results. They disclosed, for each of eight items: a “[d]escription” of the item; the weight of the item and how the weight was measured; the test(s) she performed on the item, including whether she first ran a “[b]lank” on the testing equipment; the results of those tests; and a “[c]onclusion” about the item’s identity. See id., at 88a–98a. The signed report then distilled the notes into two pages of ultimate findings, denoted “results/interpretations.” See id., at 85a–87a. After listing the eight items, the report stated that four “[c]ontained a usable quantity of methamphetamine,” three “[c]ontained a usable quantity of marijuana,” and one “[c]ontained a usable quantity of cannabis.” Id., at 86a–87a. The State originally planned for Rast to testify about those matters at Smith’s trial.

But with three weeks to go, the State called an audible, replacing Rast with a different DPS analyst as its expert witness. In the time between testing and trial, Rast had stopped working at the lab, for unexplained reasons. And the State chose not to rely on the now-former employee as a witness. So the prosecutors filed an amendment to their “final pre-trial conference statement” striking out the name Elizabeth Rast and adding “Greggory Longoni, forensic scientist (substitute expert).” Id., at 26a. Longoni had no prior connection to the Smith case, and the State did not claim otherwise. Its amendment simply stated that “Mr. Longoni will provide an independent opinion on the drug testing performed by Elizabeth Rast.” Ibid. And it continued: “Ms. Rast will not be called. [Mr. Longoni] is expected to have the same conclusion.” Ibid.

And he did come to the same conclusion, in reliance on Rast’s records. Because he had not participated in the Smith case, Longoni prepared for trial by reviewing Rast’s report and notes. And when Longoni took the stand, he referred to those materials and related what was in them, item by item by item. As to each, he described the specific “scientific method[s]” Rast had used to analyze the substance (e.g., a microscopic examination, a chemical color test, a gas chromatograph/mass spectrometer test). Id., at 41a; see id., at 42a, 46a–48a. And as to each, he stated that the testing had adhered to “general principles of chemistry,” as well as to the lab’s “policies and practices,” id., at 47a–48a; see id., at 40a; so he noted, for example, that Rast had run a “blank” to confirm that testing equipment was not contaminated, id., at 42a, 47a. After thus telling the jury what Rast’s records conveyed about her testing of the items, Longoni offered an “independent opinion” of their identity. Id., at 46a–47a, 49a. More specifically, the opinions he offered were: that Item 26 was “a usable quantity of marijuana,” that Items 20A and 20B were “usable quantit[ies] of methamphetamine,” and that Item 28 was “[a] usable quantity of cannabis.” Ibid.

After Smith was convicted, he brought an appeal focusing on Longoni’s testimony. In Smith’s view, the State’s use of a “substitute expert”—who had not participated in any of the relevant testing—violated his Confrontation Clause rights. Id., at 26a; see Brief for Appellant Smith in No. 1 CA–CR 21–0451 (Ariz. Ct. App.), pp. 20–23. The real witness against him, Smith urged, was Rast, through her written statements; but he had not had the opportunity to cross-examine her. See ibid. The State disagreed. In its view, Longoni testified about “his own independent opinions,” even though making use of Rast’s records. Brief for Appellee Arizona in No. 1 CA–CR 21–0451 (Ariz. Ct. App.), p. 22. So Longoni was the only witness Smith had a right to confront. See ibid.

The Arizona Court of Appeals affirmed Smith’s convictions, rejecting his Confrontation Clause challenge. It relied on Arizona precedent (similar to the Illinois Supreme Court’s decision in Williams) stating that an expert may testify to “the substance of a non-testifying expert’s analysis, if such evidence forms the basis of the [testifying] expert’s opinion.” App. to Pet. for Cert. 11a–12a (quoting State ex rel. Montgomery v. Karp, 236 Ariz. 120, 124, 336 P.3d 753, 757 (App. 2014)). That is because, the Arizona courts have said, the “underlying facts” are then “used only to show the basis of [the in-court witness’s] opinion and not to prove their truth.” Ibid., 336 P. 3d, at 757. On that view, the Court of Appeals held, Longoni could constitutionally “present[ ] his independent expert opinions” as “based on his review of Rast’s work.” App. to Pet. for Cert. 11a.

We granted certiorari to consider that reasoning, 600 U. S. ___ (2023), and we now reject it.[3]


Smith’s confrontation claim can succeed only if Rast’s statements came into evidence for their truth. As earlier explained, the Clause applies solely to “testimonial hearsay.” Davis, 547 U. S., at 823 (emphasis added); see supra, at 3. And that means the Clause “does not bar the use of testimonial statements for purposes other than establishing the truth of the matter asserted.” Crawford, 541 U. S., at 60, n. 9. So a court analyzing a confrontation claim must identify the role that a given out-of-court statement—here, Rast’s statements about her lab work—served at trial. On that much, indeed, the entire Williams Court agreed. Amid all the fracturing that case produced, every Justice defined its primary question in the same way: whether the absent analyst’s statements were introduced for their truth. See 567 U. S., at 57–58 (plurality opinion); id., at 104 (Thomas, J., concurring in judgment); id., at 125–126 (Kagan, J., dissenting). The parties here likewise concur in that framing. See Brief for Smith 28–29; Brief for Arizona 17–18. If Rast’s statements came in to establish the truth of what she said, then the Clause’s alarms begin to ring; but if her statements came in for another purpose, then those alarms fall quiet.

Where the parties disagree, of course, is in answering that purpose question. Smith argues that the “for the truth” condition is satisfied here, just as much as in Melendez-Diaz or Bullcoming. See Brief for Smith 23–28; supra, at 3–5. In his view, Rast’s statements were conveyed, via Longoni’s testimony, to establish that what she said happened in the lab did in fact happen. Or put more specifically, those statements were conveyed to show that she used certain standard procedures to run certain tests, which enabled identification of the seized items. The State sees the matter differently. See Brief for Arizona 19–26. Echoing the Arizona Court of Appeals (and the Illinois Supreme Court in Williams), the State argues that Rast’s statements came into evidence not for their truth, but instead to “show the basis” of the in-court expert’s independent opinion. Brief for Arizona 21; see supra, at 6. And to defend that characterization, Arizona emphasizes that its Rule of Evidence 703 (again, like Illinois’s) authorizes the admission of such statements only for that purpose—i.e., to “help[ ] the jury [to] evaluate” the opinion testimony. Brief for Arizona 20–21; see post, at 8 (Alito, J., concurring in judgment) (arguing the same as to Federal Rule of Evidence 703).

Evidentiary rules, though, do not control the inquiry into whether a statement is admitted for its truth. That inquiry, as just described, marks the scope of a federal constitutional right. See supra, at 11. And federal constitutional rights are not typically defined—expanded or contracted—by reference to non-constitutional bodies of law like evidence rules.[4] The confrontation right is no different, as Crawford made clear. “Where testimonial statements are involved,” that Court explained, “the Framers [did not mean] to leave the Sixth Amendment’s protection to the vagaries of the rules of evidence.” 541 U. S., at 61. Justice Thomas reiterated the point in Williams: “[C]oncepts central to the application of the Confrontation Clause are ultimately matters of federal constitutional law that are not dictated by state or federal evidentiary rules.” 567 U. S., at 105 (opinion concurring in judgment). We therefore do not “accept [a State’s] nonhearsay label at face value.” Id., at 106; see id., at 132 (Kagan, J., dissenting). Instead, we conduct an independent analysis of whether an out-of-court statement was admitted for its truth, and therefore may have compromised a defendant’s right of confrontation.

We did just that in Tennessee v. Street—and in so doing showcased how an out-of-court statement can come into evidence for a non-truth-related reason. See 471 U. S., at 410–417. Street was charged with murder, based mostly on a stationhouse confession. At trial, he claimed that the confession was coerced, and in a peculiar way: The sheriff, he said, had read aloud an accomplice’s confession and forced him to repeat it. On rebuttal, the State introduced the other confession (through the sheriff ’s testimony) to demonstrate to the jury all the ways its content deviated from Street’s. We upheld that use as “nonhearsay.” Id., at 413. The other confession came in, we explained, not to prove “the truth of [the accomplice’s] assertions” about how the murder happened, but only to disprove Street’s claim about how the sheriff elicited his own confession. Ibid. Or otherwise said, the point was to show, by highlighting the two confessions’ differences, that Street’s was not a “coerced imitation.” Id., at 414. For that purpose, the truth of the accomplice’s confession (and the credibility of the accomplice himself ) was irrelevant.

But truth is everything when it comes to the kind of basis testimony presented here. If an expert for the prosecution conveys an out-of-court statement in support of his opinion, and the statement supports that opinion only if true, then the statement has been offered for the truth of what it asserts. How could it be otherwise? “The whole point” of the prosecutor’s eliciting such a statement is “to establish—because of the [statement’s] truth—a basis for the jury to credit the testifying expert’s” opinion. Stuart, 586 U. S., at ___ (Gorsuch, J., dissenting from denial of certiorari) (slip op., at 3) (emphasis in original). Or said a bit differently, the truth of the basis testimony is what makes it useful to the prosecutor; that is what supplies the predicate for—and thus gives value to—the state expert’s opinion. So “[t]here is no meaningful distinction between disclosing an out-of-court statement” to “explain the basis of an expert’s opinion” and “disclosing that statement for its truth.” Williams, 567 U. S., at 106 (Thomas, J., concurring in judgment). A State may use only the former label, but in all respects the two purposes merge.

Or to see the point another way, consider it from the factfinder’s perspective. In the view of the Arizona courts, an expert’s conveyance of another analyst’s report enables the factfinder to “determine whether [the expert’s] opinion should be found credible.” Karp, 236 Ariz., at 124, 336 P. 3d, at 757; see Williams, 238 Ill. 2d, at 144, 939 N. E. 2d, at 278 (also stating that such a report “aid[s] the jury in assessing the value of [the expert’s] opinion”); supra, at 6, 10. That is no doubt right. The jury cannot decide whether the expert’s opinion is credible without evaluating the truth of the factual assertions on which it is based. See D. Kaye, D. Bernstein, A. Ferguson, M. Wittlin, & J. Mnookin, The New Wigmore: Expert Evidence §5.4.1, p. 271 (3d ed. 2021). If believed true, that basis evidence will lead the jury to credit the opinion; if believed false, it will do the opposite. See Williams, 567 U. S., at 106, and n. 1 (Thomas, J., concurring in judgment); id., at 126–127 (Kagan, J., dissenting). But that very fact is what raises the Confrontation Clause problem. For the defendant has no opportunity to challenge the veracity of the out-of-court assertions that are doing much of the work.

And if that explanation seems a bit abstract, then take this case as its almost-too-perfect illustration. Recall that Rast tested eight seized items before she disappeared from the scene. At trial, the prosecutor asked the State’s “substitute expert” Longoni to testify about four of them (with the rest dropping out of the case). App. to Pet. for Cert. 26a. A recap of their exchange about one item will be enough; the rest followed the same pattern. Remember as you read that Longoni, though familiar with the lab’s general practices, had no personal knowledge about Rast’s testing of the seized items. Rather, as his testimony makes clear, what he knew on that score came only from reviewing Rast’s records. With that as background:

Q Turn your attention to Item 26. I’m going to hand you what’s been marked as State’s Exhibit 98 [Rast’s notes]. . . . Did you review how [Item] 26 was tested in this case?

A Yes.

Q When you reviewed it, did you notice whether the [standard lab] policies and practices that you have just described were followed?

A Yes.

Q Were they followed?

A Yes.

.     .     .     .     .

Q From your review of the lab notes in this case, can you tell me what scientific method was used to analyze Item 26?

A Yes.

Q And what was used?

A The microscopic examination and the chemical color test. . . .

Q That was done in this case?

A Yes, it was.

Q Was there a blank done to prevent contamination, make sure everything was clean?

A According to the notes, yes.

.     .     .     .     .

Q In reviewing what was done, your knowledge and training as a forensic scientist, your knowledge and experience with DPS’s policies, practices, procedures, your knowledge of chemistry, the lab notes, the intake records, the chemicals used, the tests done, can you form an independent opinion on the identity of Item 26?

A Yes.

Q What is that opinion?

A That is a usable quantity of marijuana.

Id., at 39a–42a, 46a. And then the prosecutor went on to Items 20A, 20B, and 28, asking similar questions, receiving similar answers based on Rast’s records, and finally eliciting similar “independent opinions”—which were no more than what Rast herself had concluded. See supra, at 8–9. “Yes,” Longoni confirmed, just as Item 26 was a “usable quantity of marijuana,” Items 20A and 20B were “usable quantit[ies] of methamphetamine” and Item 28 was a “usable quantity of cannabis.” App. to Pet. for Cert. 46a, 47a, 49a.

Rast’s statements thus came in for their truth, and no less because they were admitted to show the basis of Longoni’s expert opinions. All those opinions were predicated on the truth of Rast’s factual statements. Longoni could opine that the tested substances were marijuana, methamphetamine, and cannabis only because he accepted the truth of what Rast had reported about her work in the lab—that she had performed certain tests according to certain protocols and gotten certain results. And likewise, the jury could credit Longoni’s opinions identifying the substances only because it too accepted the truth of what Rast reported about her lab work (as conveyed by Longoni). If Rast had lied about all those matters, Longoni’s expert opinion would have counted for nothing, and the jury would have been in no position to convict. So the State’s basis evidence—more precisely, the truth of the statements on which its expert relied—propped up its whole case. But the maker of those statements was not in the courtroom, and Smith could not ask her any questions.

Approving that practice would make our decisions in Melendez-Diaz and Bullcoming a dead letter, and allow for easy evasion of the Confrontation Clause. As earlier described, those two decisions applied Crawford in “straightforward” fashion to forensic evidence. Melendez-Diaz, 557 U. S., at 312; see Bullcoming, 564 U. S., at 659–661; supra, at 3–5. The first prevented the introduction of a lab analyst’s testimonial report sans lab analyst. The second refused to accede to the idea that any old analyst—i.e., a substitute who had not taken part in the lab work—would do. Arizona offers only a slight variation. On its view, a surrogate analyst can testify to all the same substance—that is, someone else’s substance—as long as he bases an “independent opinion” on that material. And that is true even if, as here, the proffered opinion merely replicates, rather than somehow builds on, the testing analyst’s conclusions. So every testimonial lab report could come into evidence through any trained surrogate, however remote from the case. And no defendant would have a right to cross-examine the testing analyst about what she did and how she did it and whether her results should be trusted. In short, Arizona wants to end run all we have held the Confrontation Clause to require. It cannot.

Properly understood, the Clause still allows forensic experts like Longoni to play a useful role in criminal trials. Because Longoni worked in the same lab as Rast, he could testify from personal knowledge about how that lab typically functioned—the standards, practices, and procedures it used to test seized substances, as well as the way it maintained chains of custody. (Indeed, Longoni did just that in a different part of his testimony. See App. to Pet. for Cert. 32a–39a.) Or had he not been familiar with Rast’s lab, he could have testified in general terms about forensic guidelines and techniques—perhaps explaining what it means for a lab to be accredited and what requirements accreditation imposes. Or as the Williams plurality and dissent both observed, he might have been asked—and could have answered—any number of hypothetical questions, taking the form of: “If or assuming some out-of-court statement were true, what would follow from it?” See 567 U. S., at 67–68; id., at 129, n. 2. (The State of course would then have to separately prove the thing assumed.) The United States, appearing as amicus curiae in support of neither party, usefully addressed these matters at oral argument, distinguishing Longoni’s testimony as block-quoted above from the various kinds of testimony just described. See Tr. of Oral Arg. 36–41. The latter forms of testimony allow forensic expertise to inform a criminal case without violating the defendant’s right of confrontation. And we offer these merely as examples; there may be others.

But as the United States acknowledged, the bulk of Longoni’s testimony took no such permissible form. Ibid. Here, the State used Longoni to relay what Rast wrote down about how she identified the seized substances. Longoni thus effectively became Rast’s mouthpiece. He testified to the precautions (she said) she took, the standards (she said) she followed, the tests (she said) she performed, and the results (she said) she obtained. The State offered up that evidence so the jury would believe it—in other words, for its truth. So if the out-of-court statements were also testimonial, their admission violated the Confrontation Clause. Smith would then have had a right to confront the person who actually did the lab work, not a surrogate merely reading from her records.


What remains is whether the out-of-court statements Longoni conveyed were testimonial. As earlier explained, that question is independent of everything said above: To implicate the Confrontation Clause, a statement must be hearsay (“for the truth”) and it must be testimonial—and those two issues are separate from each other. See supra, at 3. The latter, this Court has stated, focuses on the “primary purpose” of the statement, and in particular on how it relates to a future criminal proceeding. See ibid. (noting varied formulations of the standard).[5] A court must therefore identify the out-of-court statement introduced, and must determine, given all the “relevant circumstances,” the principal reason it was made. Bryant, 562 U. S., at 369.

But that issue is not now fit for our resolution. The question presented in Smith’s petition for certiorari did not ask whether Rast’s out-of-court statements were testimonial. See supra, at 11, n. 3 (quoting Pet. for Cert. i). Instead, it took as a given that they were. See id., at i. That presentation reflected the Arizona Court of Appeals’ opinion. As described earlier, that court relied on the “not for the truth” rationale we have just rejected. See supra, at 10. It did not decide whether Rast’s statements were testimonial. Nor, to our knowledge, did the trial court ever take a stance on that issue. Because “we are a court of review, not of first view,” we will not be the pioneer court to decide the matter. Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005). And indeed, we are not sure if there remains a matter to decide. Smith argues that the State has forfeited the argument: Arizona, he says, “gave no hint in the proceedings below that it believed Rast’s statements were anything but testimonial.” Reply Brief 3. The State denies that assertion, pointing to a passage about Williams in its lower court briefing. See Brief for Arizona 39, n. 14. The dispute is best addressed by a state court. So we return the testimonial issue, including the threshold forfeiture matter, to the Arizona Court of Appeals.

But we offer a few thoughts, based on the arguments made here, about the questions the state court might usefully address if the testimonial issue remains live. First, the court will need to consider exactly which of Rast’s statements are at issue. In this Court, the parties disputed whether Longoni was reciting from Rast’s notes alone, or from both her notes and final report. See supra, at 8–9 (describing those documents). In Arizona’s view, everything Longoni testified to came from Rast’s notes; although he at times used the word “report,” a close comparison of the documents and his testimony reveals (the State says) that he meant only the notes. See Brief for Arizona 39–40; Tr. of Oral Arg. 69–72; see also App. to Pet. for Cert. 39a–40a, 48a. Smith disagrees, taking Longoni’s references to the “report,” as well as the notes, at face value. According to Smith, Longoni “relied on both” documents and in fact “treated them as a unit,” with the notes “attached” to the report as “essentially an appendix.” Reply Brief 4; Tr. of Oral Arg. 25, 98. Resolving that dispute might, or then again might not, affect the court’s ultimate disposition of Smith’s Confrontation Clause claim. We note only that before the court can decide the primary purpose of the out-of-court statements introduced at Smith’s trial, it needs to determine exactly what those statements were.

In then addressing the statements’ primary purpose—why Rast created the report or notes—the court should consider the range of recordkeeping activities that lab analysts engage in. See generally supra, at 3 (describing formulations of the testimonial inquiry). After all, some records of lab analysts will not have an evidentiary purpose. The United States as amicus curiae notes, for example, that lab records may come into being primarily to comply with laboratory accreditation requirements or to facilitate internal review and quality control. See Tr. of Oral Arg. 51. Or some analysts’ notes may be written simply as reminders to self. See id., at 20, 52. In those cases, the record would not count as testimonial. To do so, the document’s primary purpose must have “a focus on court.” Id., at 52. And again, the state court on remand should make that assessment as to each record whose substance Longoni conveyed.


Our holding today follows from all this Court has held about the Confrontation Clause’s application to forensic evidence. A State may not introduce the testimonial out-of-court statements of a forensic analyst at trial, unless she is unavailable and the defendant has had a prior chance to cross-examine her. See Crawford, 541 U. S., at 68; Melendez-Diaz, 557 U. S., at 311. Neither may the State introduce those statements through a surrogate analyst who did not participate in their creation. See Bullcoming, 564 U. S., at 663. And nothing changes if the surrogate—as in this case—presents the out-of-court statements as the basis for his expert opinion. Those statements, as we have explained, come into evidence for their truth—because only if true can they provide a reason to credit the substitute expert. So a defendant has the right to cross-examine the person who made them.

That means Arizona does not escape the Confrontation Clause just because Rast’s records came in to explain the basis of Longoni’s opinion. The Arizona Court of Appeals thought otherwise, and so we vacate its judgment. To address the additional issue of whether Rast’s records were testimonial (including whether that issue was forfeited), we remand the case for further proceedings not inconsistent with this opinion.

It is so ordered.

1  The Court also failed to reach agreement on the testimonial issue. The four Justices who accepted the state court’s “not for the truth” view also concluded that the report was not testimonial. See Williams, 567 U. S., at 81–86 (plurality opinion). But they did so for reasons different from Justice Thomas’s. Compare ibid. with id., at 110–117 (opinion concurring in judgment). The result was that no single rationale for affirmance garnered a majority.
2  Compare, e.g., State v. Mercier, 2014 ME 28, ¶¶12–14, 87 A.3d 700, 704 (accepting the “not for the truth” rationale for admitting an expert’s basis testimony); State v. Hutchison, 482 S.W.3d 893, 914 (Tenn. 2016); United States v. Murray, 540 Fed. Appx. 918, 921 (CA11 2013), with People v. Sanchez, 63 Cal. 4th 665, 684, 374 P.3d 320, 333 (2016) (rejecting the “not for the truth” rationale for admitting an expert’s basis testimony); Martin v. State, 60 A.3d 1100, 1107 (Del. 2013); Young v. United States, 63 A.3d 1033, 1045 (D. C. 2013); Leidig v. State, 475 Md. 181, 234, n. 23, 256 A.3d 870, 901, n. 23 (2021); Commonwealth v. Jones, 472 Mass. 707, 714, 37 N. E. 3d 589, 597 (2015).
3  The question on which we granted certiorari made reference as well to another aspect of the Court of Appeals’ reasoning. That question asks whether the Confrontation Clause permits “testimony by a substitute expert conveying the testimonial statements of a nontestifying forensic analyst, on the grounds that (a) the testifying expert offers some independent opinion and the analyst’s statements are offered not for their truth but to explain the expert’s opinion, and (b) the defendant did not independently seek to subpoena the analyst.” Pet. for Cert. i. The “(b)” in that question arises from the following sentence in the court’s opinion: “Had Smith sought to challenge Rast’s analysis, he could have called her to the stand and questioned her, but he chose not to do so.” App. to Pet. for Cert. 12a. We need not spend much time on that rationale because the State rightly does not defend it. As we held in Melendez-Diaz, a defendant’s “ability to subpoena” an absent analyst “is no substitute for the right of confrontation.” 557 U. S., at 324. The Confrontation Clause “imposes a burden on the prosecution to present its witnesses, not on the defendant to bring those adverse witnesses into court.” Ibid.
4  One qualification is appropriate. If an evidentiary rule reflects a long-established understanding, then it might shed light on the historical meaning of the Confrontation Clause. But that could not possibly be said of Rule 703—the rule Arizona cites to support the introduction of basis evidence. On the contrary, that rule is a product of the late-20th century, and was understood from the start to depart from past practice. See Brief for Richard D. Friedman as Amicus Curiae 17; Advisory Committee’s Notes on Fed. Rule Evid. 703, 28 U. S. C. App., p. 393.
5  Given that focus, the mine-run of materials on which most expert witnesses rely in forming opinions—including books and journals, surveys, and economic or scientific studies—will raise no serious confrontation issues. See Brief for United States as Amicus Curiae 13–17 (giving examples of classic expert-basis evidence). That is because the preparation of those materials generally lacks any “evidentiary purpose.” Melendez-Diaz, 557 U. S., at 311.


No. 22–899



on writ of certiorari to the court of appeals of arizona, division one

[June 21, 2024]

Justice Thomas, concurring in part.

I join the Court in all but Part III of its opinion. The Sixth Amendment’s Confrontation Clause provides: “In all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.” This Clause bars the admission of an absent witness’s testimonial statements for their truth, unless the witness is unavailable and the defendant previously had an opportunity to cross-examine that witness. See Crawford v. Washington, 541 U.S. 36, 50–56, 60, n. 9 (2004). Today, the Court correctly concludes that “[w]hen an expert conveys an absent analyst’s statements in support of his opinion, and the statements provide that support only if true, then the statements come into evidence for their truth.” Ante, at 2; see also Williams v. Illinois, 567 U.S. 50, 106 (2012) (Thomas, J., concurring in judgment). But, a question remains whether that analyst’s statements were testimonial. I agree with the Court that, because the courts below did not consider this question, we should remand for the Arizona Court of Appeals to answer it in the first instance. Ante, at 19–20. But, I disagree with the Court’s suggestion that the Arizona Court of Appeals should answer that question by looking to each statement’s “primary purpose.” Ante, at 20–21.

I continue to adhere to my view that “the Confrontation Clause is implicated by extrajudicial statements only insofar as they are contained in formalized testimonial materials, such as affidavits, depositions, prior testimony, or confessions.” [1]* White v. Illinois, 502 U.S. 346, 365 (1992) (Thomas, J., concurring in part and concurring in judgment); see also Ohio v. Clark, 576 U.S. 237, 254–255 (2015) (Thomas, J., concurring in judgment); Williams, 567 U. S., at 110–111 (opinion of Thomas, J.); Michigan v. Bryant, 562 U.S. 344, 379 (2011) (Thomas, J., concurring in judgment); Melendez-Diaz v. Massachusetts, 557 U.S. 305, 329 (2009) (Thomas, J., concurring); Giles v. California, 554 U.S. 353, 377–378 (2008) (Thomas, J., concurring); Davis v. Washington, 547 U.S. 813, 837 (2006) (opinion of Thomas, J.); Lilly v. Virginia, 527 U.S. 116, 143 (1999) (Thomas, J., concurring in part and concurring in judgment). The Confrontation Clause guarantees a criminal defendant “the right . . . to be confronted with the witnesses against him.” Amdt. 6. As I have previously explained, “[w]itnesses . . . are those who bear testimony. And testimony is a solemn declaration or affirmation made for the purpose of establishing or proving some fact.” Davis, 547 U. S., at 836 (opinion of Thomas, J.) (citation, internal quotation marks, and alterations omitted). This understanding is grounded in “[t]he history surrounding the right to confrontation,” which “was developed to target particular practices that occurred under the English bail and committal statutes passed during the reign of Queen Mary, namely, the civil-law mode of criminal procedure, and particularly its use of ex parte examinations as evidence against the accused.” Id., at 835 (internal quotation marks omitted). Rather than attempt to divine a statement’s “primary purpose,” I would look for whether the statement is “similar in solemnity to the Marian examination practices that the Confrontation Clause was designed to prevent.” Williams, 567 U. S., at 112 (opinion of Thomas, J.). In my view, the Arizona Court of Appeals should consider on remand whether the statements at issue have the requisite formality and solemnity to qualify as testimonial. If they do not, the Confrontation Clause poses no barrier to their admission.

1 *The Confrontation Clause “also reaches the use of technically informal statements when used to evade the formalized process.” Davis v. Washington, 547 U.S. 813, 838 (2006) (Thomas, J., concurring in judgment in part and dissenting in part).


No. 22–899



on writ of certiorari to the court of appeals of arizona, division one

[June 21, 2024]

Justice Gorsuch, concurring in part.

I am pleased to join the Court’s opinion holding that, when an expert presents another’s statements as the “basis” for his own opinion, he is offering those statements for their truth. See Parts I, II, and IV, ante.

I cannot join, however, the Court’s discussion in Part III about when an absent analyst’s statement might qualify as “testimonial.” See ante, at 19–21. As the Court says, “that issue is not now fit for our resolution.” Ante, at 19. It was not part of the question presented for our review, nor was it the focus of the decision below. Ante, at 19–20. In fact, the State devoted so little attention to the “testimonial” issue in the Arizona courts that any argument it might make on the subject on remand may be forfeited. Ante, at 20. Further, the Court’s thoughts on the subject are in no way necessary to the resolution of today’s dispute. What makes a statement testimonial, the Court notes, is an entirely “separate” issue. Ante, at 19.

Nor am I entirely sure about the guidance found in Part III. The Sixth Amendment protects the accused’s “right . . . to be confronted with the witnesses against him.” As the Court sees it, whether a statement being offered for its truth and tendency to inculpate a defendant triggers that right depends “on the ‘primary purpose’ of the statement, and in particular on how it relates to a future criminal proceeding.” Ante, at 19. I cannot help but wonder whether that is correct.

Just consider a few other possibilities. In protecting the right to confront “witnesses,” perhaps the Sixth Amendment reaches any “person who gives or furnishes evidence.” United States v. Hubbell, 530 U.S. 27, 49–50 (2000) (Thomas, J., concurring) (discussing founding-era meaning of the word “witness” in the Fifth Amendment); see also id., at 50, n. 1. Or perhaps the Amendment reaches all “those who ‘bear testimony.’ ” Crawford v. Washington, 541 U.S. 36, 51 (2004) (quoting 2 N. Webster, An American Dictionary of the English Language (1828)). Perhaps, too, a statement “bears testimony” so long as it “explicitly or implicitly . . . relate[s] a factual assertion or disclose[s] information.” Doe v. United States, 487 U.S. 201, 210 (1988) (discussing what makes a statement “testimonial” for purposes of the Fifth Amendment); see also 2 Webster, An American Dictionary (observing near the founding that “testimony” could mean “evidence” and “proof of some fact” as well as a “solemn declaration or affirmation” made to “establis[h] or prov[e] some fact”). To my mind, all these questions (and maybe others too) warrant careful exploration in a case that presents them and, without more assurance, I worry that the Court’s proposed “primary purpose” test may be a limitation of our own creation on the confrontation right.

I am concerned, as well, about the confusion a “primary purpose” test may engender. Does it focus, for example, on the purposes an objective observer would assign to a challenged statement, see ante, at 3 (referencing the “ ‘objective witness’ ”), the declarant’s purposes in making it, see ante, at 21 (asking “why Rast created the report or notes”), the government’s purposes in “ ‘procur[ing]’ ” it, see ante, at 3, or maybe still some other point of reference? Even after we figure out a statement’s purposes, how do we pick the primary one out of the several a statement might serve? Or determine in exactly what way that purpose must “relat[e] to a future criminal proceeding”? Ante, at 19. And if we fail to find some foothold in text and historical practice for resolving these questions, how can judges answer them without resort to their own notions of what would be best?

Some time ago, Chief Justice Marshall charged the judiciary with “be[ing] watchful of every inroad” on the accused’s right to be confronted with the witnesses against him. United States v. Burr, 25 F. Cas. 187, 193 (No. 14,694) (CC Va. 1807). With that cautionary note in mind, I respectfully concur in all but Part III of the Court’s opinion.



No. 22–899



on writ of certiorari to the court of appeals of arizona, division one

[June 21, 2024]

Justice Alito, with whom the Chief Justice joins, concurring in the judgment.

Today, the Court inflicts a needless, unwarranted, and crippling wound on modern evidence law. There was a time when expert witnesses were required to express their opinions as responses to hypothetical questions. But eventually, this highly artificial, awkward, confusing, and abuse-laden form of testimony earned virtually unanimous condemnation. More than a century ago, judges, evidence scholars, and legal reform associations began to recommend that courts abandon the required use of hypotheticals, and more than 50 years ago, the Federal Rules of Evidence did so. Now, however, the Court proclaims that a prosecution expert will frequently violate the Confrontation Clause when he testifies in strict compliance with the Federal Rules of Evidence and similar modern state rules. Instead, the Court suggests that such experts revert to the form that was buried a half-century ago. Ante, at 18. There is no good reason for this radical change.


To explain why I think the Court has gone far astray, I begin with a brief look at the history of expert testimony—and particularly, why the hypothetical-question requirement was replaced by the (superior) mode of testimony allowed by the Federal Rules of Evidence.


Expert testimony presents a challenge for a legal system like ours that restricts a fact-finder’s ability to consider hearsay. This is so because an expert’s opinion very often is based on facts that are not proved in court. As a modern treatise puts it, the value of experts lies in their ability to “brin[g] to bear a body of knowledge largely extraneous to the facts of the particular case.” D. Kaye, D. Bernstein, A. Ferguson, M. Wittlin, & J. Mnookin, The New Wigmore: Expert Evidence §1.2.1, p. 4 (3d ed. 2021) (Kaye). Wigmore made the same point when he wrote that “[n]o one professional man can know from personal observation more than a minute fraction of the data which he must every day treat as working truths.” 1 J. Wigmore, Evidence §665(3), p. 762 (1904) (Wigmore). Instead, experts routinely “rel[y] on the reported data of fellow-scientists, learned by perusing their reports in books and journals.” Id., at 762–763 (emphasis deleted); see also Kaye §4.1, at 165 (“[P]art of an expert’s very expertise inevitably derive[s] from hearsay”).

Despite this problem, courts in Great Britain and this country long ago recognized the value of expert testimony and concluded that they “must . . . accept this kind of knowledge from scientific men,” even if it meant allowing testimony based on facts of which the expert did not have firsthand knowledge. See 1 Wigmore 763; 1 S. Greenleaf, Evidence §430(l), p. 529 (rev. 16th ed. 1899) (“It would be absurd to deny judicial standing to such knowledge, because all scientific data must be handed down from generation to generation by hearsay, and each student can hope to test only a trifling fraction of scientific truth by personal experience”); Slocovich v. Orient Mut. Ins. Co., 108 N.Y. 56, 64, 14 N.E. 802, 805 (1888) (“An expert is qualified to give evidence as to things which he has never seen. He may base an opinion upon facts proved by other witnesses, or upon facts assumed and embraced within the case”).

Recognizing this reality, a court in the late-18th century admitted expert testimony about the seaworthiness of a ship based on a survey conducted when the expert was not present. Thornton v. Royal Exchange Assurance Co., Peake 37, 38, 170 Eng. Rep. 70, 71 (N. P. 1790). Similarly, an early-19th century decision allowed ship surveyors to testify to the seaworthiness of a vessel they had never seen. Beckwith v. Sydebotham, 1 Camp. 116, 170 Eng. Rep. 897 (N. P. 1807). The opposing party objected that the experts did not know the underlying facts to be true, but the court admitted their opinions because the experts’ technical knowledge could assist the jury. Ibid. The fact that “the truth of the facts stated to them was not certainly known” went to the weight of the testimony, not its admissibility. Ibid.

Throughout the 19th and into the 20th century, experts generally testified in the form of an opinion in response to a hypothetical question. An attorney would ask an expert to assume that certain facts were true and would then query whether a particular conclusion could conceivably follow. See 3 S. Saltzburg, M. Martin, D. Capra, & J. Berch, Federal Rules of Evidence Manual §703.02[1] (13th ed. 2023).

This procedure was highly artificial because it bore little resemblance to the way in which experts actually form opinions. And the procedure surely did not conform to the way lay jurors think and speak.

The procedure’s aim was to prevent a jury from jumping to the conclusion that the facts packed into the hypothetical were true, but it is questionable whether the practice achieved that objective. For instance, here is the question that defense counsel asked a psychiatric witness in Charles Guiteau’s trial for murdering President Garfield:

“Q. . . . Assume it to be a fact that there was a strong hereditary taint of insanity in the blood of the prisoner at the bar; also that at about the age of thirty-five years his mind was so much deranged that he was a fit subject to be sent to an insane asylum; also that at different times from that date during the next succeeding five years he manifested such decided symptoms of insanity, without stimulation, that many different persons conversing with him and observing his conduct believed him to be insane; also that during the month of June, 1881, at about the expiration of said term of five years, he honestly became dominated by the idea that he was inspired of God to remove by death the President of the United States; also that he acted upon what he believed to be such inspiration, and what he believed to be in accordance with the Divine will, in preparation for and in the accomplishment of such purpose; also that he committed the act of shooting the President under what he believed to be a Divine command which he was not at liberty to disobey, and which belief amounted to a conviction that controlled his conscience and over-powered his will as to that act, so that he could not resist the mental pressure upon him; also that immediately after the shooting he appeared calm and as one relieved by the performance of a great duty; also that there was no other adequate motive for the act than the conviction that he was executing the Divine will for the good of his country—assuming all these propositions to be true, state whether in your opinion the prisoner was sane or insane at the time of shooting President Garfield?

“A. Assuming those to be true, I should say the prisoner was insane.” C. Rosenberg, The Trial of the Assassin Guiteau 144–145 (1968) (Rosenberg).

How likely is it that a jury hearing a question like that would keep in mind that all the facts loaded into the question were merely hypothetical and not necessarily supported by the evidence in the case?

The Guiteau example illustrates many other problems with hypothetical questioning. For one, hypothetical questions were “difficult for the attorneys to frame, for the court to rule on, and for the jury to understand.” M. Ladd, Expert Testimony, 5 Vand. L. Rev. 414, 425 (1952) (Ladd). Like the question above, the hypotheticals were often “so built up and contrived” that they were impossible for either the jury or the expert to follow. 1 J. Wigmore, Evidence 1095 (2d ed. 1923) (1 Wigmore 2d); accord, Ladd 427. One case involved a hypothetical that extended over “eighty-three pages of typewritten transcript, and an objection involved in fourteen pages more of the record.” Treadwell v. Nickel, 194 Cal. 243, 266, 228 P. 25, 35 (1924). Such questions required an expert to have the extraordinary ability “to comprehend in one mental operation the entirety of what has been asked so as to give any answer.” Ladd 427; see, e.g., Editorials, The Hypothetical Question Again, 24 J. Crim. L. & C. 517, 517–519 (1933). And juries surely found following lengthy hypotheticals even more mystifying.

For another, lawyers often used hypotheticals as a preview of their closing arguments. See, e.g., Rosenberg 144 (“Assume . . . that he committed the act of shooting the President under what he believed to be a Divine command which he was not at liberty to disobey . . . so that he could not resist the mental pressure upon him”); see also S. Gross, Expert Evidence, 1991 Wis. L. Rev. 1113, 1162 (Gross); Wigmore 2d §686, at 1095; Ladd 246. In doing so, they sometimes sneaked in “irrelevant” information, Gross 1162, and excluded necessary details, W. White, Insanity and the Criminal Law 86 (1923) (White) (describing the hypothetical question as “eliminat[ing] from consideration every human element which every common-sense man takes into consideration when he formulates an opinion”). One medical expert declared that he “ha[d] never known a hypothetical question, in a trial involving the mental condition of the defendant, which in [his] opinion offered a fair presentation of the case.” Ibid. As a result, experts either provided answers that were entirely disconnected from “the actual case,” 1 Wigmore 2d §686, at 1095, or else they ignored the hypothetical altogether, White 87.

Because opposing counsel often disagreed for strategic reasons about which facts should be included in a hypothetical, constructing a hypothetical that the judge would permit was often a tricky and contentious business. If counsel did not include enough facts to satisfy opposing counsel, the hypothetical would be met with an objection, and its sufficiency would provide grist for an appeal. F. Rossi, Expert Witnesses 114 (1991). The threat of dragging out litigation led counsel to make their hypotheticals even longer and more confusing. Ibid.

By the early-20th century, this form of testimony was scorned. In the second edition of his treatise, issued in 1923, Wigmore proclaimed the hypothetical question “that feature which does most to disgust men of science with the law of Evidence.” 1 Wigmore 2d §686, at 1094. Around the same time, Judge Learned Hand labeled hypotheticals “the most horific and grotesque wen upon the fair face of justice.” Address of L. Hand: The Deficiencies of Trials to Reach the Heart of the Matter, in Lectures on Legal Topics, 1921–1922, p. 104 (1926). Professor Charles T. McCormick described hypotheticals as “an obstruction to the administration of justice.” Some Observations Upon the Opinion Rule and Expert Testimony, 23 Texas L. Rev. 109, 128 (1945) (McCormick). Experts shared these concerns; one lamented that lawyers’ use of hypothetical questions was often “so unfair and confusing and degrading that it does not clarify the issue nor help achieve justice.” H. Hulbert, Psychiatric Testimony in Probate Proceedings, 2 Law & Contemp. Prob. 448, 455 (1935). Eventually, the use of hypothetical questions was “nearly universally recognized as a practical disaster” by lawyers, judges, and witnesses alike. Kaye §4.4, at 189.

This state of affairs sparked efforts to eliminate hypothetical questions as a requirement. See, e.g., 1 Wigmore 2d §686, at 1094 (“The Hypothetical Question must go, as a requirement. Its abuses have become so obstructive and nauseous that no remedy short of extirpation will suffice” (emphasis deleted)). Change began first in the courts, which allowed experts to sit through trial and then provide their opinion “ ‘upon the evidence.’ ” 3 C. Chamberlayne, Modern Law of Evidence §§2482, 2483, pp. 3343–3346 (1912).

More formalized rule changes soon followed. In 1937, the Commissioners on Uniform State Laws incorporated a provision in their Model Expert Testimony Act that permitted experts to give their opinions without preliminarily disclosing their underlying facts or data. Advisory Committee’s Notes to Fed. Rule Evid. 705. In quick succession, both the Model Code of Evidence, issued by the American Law Institute in 1942, and the Uniform Rules of Evidence, approved by the American Bar Association in 1953, recommended abandonment of hypothetical questions. See ALI, Model Code of Evidence Rule 409, Comment b, p. 211 (the hypothetical question “has been so grossly abused as to be almost a scandal”); Uniform Rules of Evidence, Rule 58, Comment, p. 194 (“This rule does away with the necessity of following the practice (grossly abused) of using the hypothetical question”). In 1972, the Federal Rules of Evidence followed suit with Rules 703 and 705, and many States made similar changes.


What replaced hypotheticals was the procedure exemplified by the Federal Rules of Evidence.[1] Rule 703 provides that an expert’s opinion may be based on “facts or data in the case that the expert has been made aware of or personally observed.” And “[u]nless the court orders otherwise,” Rule 705 permits the expert to “state an opinion—and give the reasons for it—without first testifying to the underlying facts or data.”

These facts or data need not be “admissible” in evidence, and they are not admitted for the truth of what they assert. Fed. Rule Evid. 703. Instead, these facts or data may, under some circumstances, be disclosed to the jury for a limited purpose: to assist the jurors in judging the weight that should be given to the expert’s opinion. Ibid. However, this is not allowed unless the court determines that “their probative value in helping the jury evaluate the [expert’s] opinion substantially outweighs their prejudicial effect.” Ibid. And to prevent the jury from improperly relying on basis testimony for the truth of the matters it asserts, a judge must instruct the jury upon request to consider such evidence only to assess the quality of the expert’s testimony (i.e., to determine whether an expert’s statements are reliable). See Advisory Committee’s Notes on Fed. Rule Evid. 703, 28 U. S. C. App., p. 393; Fed. Rule Evid. 105 (“If the court admits evidence that is admissible . . . for a [limited] purpose—but not . . . for another purpose—the court, on timely request, must restrict the evidence to its proper scope and instruct the jury accordingly”).

This procedure is sensitive to the risk of jurors’ mistakenly treating an expert’s basis testimony as evidence of the truth of the facts of data upon which the expert relied. The Rules provide important safeguards against this danger, such as the stringent “probative value versus potential prejudice” test and the requirement that a limiting instruction be given upon request. Plus, of course, an expert’s lack of personal knowledge of the “facts or data” that are called to his attention can be brought out in cross examination and stressed in a closing argument.

This modern system is more honest because it reflects how experts actually form opinions. See Advisory Committee’s Notes on Fed. Rule Evid. 703, at 393 (describing the Rule as “designed to . . . bring the judicial practice in line with the practice of the experts themselves when not in court”). It is simpler and less likely to confuse. And it avoids many of the pitfalls of the old procedure. It may not be perfect—and evidence scholars have proposed a variety of reforms—but it is unquestionably better than the old regime it replaced.


In light of the woeful history of expert testimony by hypotheticals, why has the Court disinterred that procedural monstrosity? The Court reasons that “[i]f an expert for the prosecution conveys an out-of-court statement in support of his opinion, and the statement supports that opinion only if true, then the statement has been offered for the truth of what it asserts.” Ante, at 14. Or put differently, “the truth of the basis testimony is what makes it useful to the prosecutor; that is what supplies the predicate for—and thus gives value to—the state expert’s opinion.” Ibid. In other words, the Court seems to think that all basis testimony is necessarily offered for its truth.

This is just plain wrong. What makes basis evidence “useful” is the assistance it gives the fact-finder in judging the weight that should be given to the expert’s opinion. See Advisory Committee’s Notes on Rule 703 (basis testimony may be brought before a jury to help it “evaluate the . . . opinion”). And a trial judge must, upon request, instruct the jury to consider it only for that purpose. If a judge rules that basis evidence is not admitted for its truth and so instructs the jury, where does the Court discern a Confrontation Clause problem?

The only possible explanation is that the Court believes that juries are incapable of following such an instruction, but that conclusion is inconsistent with commonplace trial practice and with a whole string of our decisions. It is a routine matter for trial judges to instruct juries that evidence is admitted for only a limited purpose. This Court acknowledged as much in United States v. Abel, 469 U.S. 45 (1984), when it noted that “there is no rule of evidence which provides that testimony admissible for one purpose and inadmissible for another purpose is thereby rendered inadmissible; quite the contrary is the case.” Id., at 56. In such instances, courts use limiting instructions. See Fed. Rule Evid. 105; 1 R. Mosteller et al., McCormick on Evidence §59, pp. 481–483 (8th ed. 2020).

And this Court has repeatedly upheld that practice—even in “situations with potentially life-and-death stakes for defendants” and even with respect to statements that are “some of the most compelling evidence of guilt available to a jury,” Samia v. United States, 599 U.S. 635, 646–647 (2023). These decisions “credi[t] jurors by refusing to assume that they are either ‘too ignorant to comprehend, or were too unmindful of their duty to respect, instructions’ of the court.” Id., at 647. Indeed, we have described the assumption “ ‘that juries will follow the instructions given them by the trial judge’ ” as “ ‘crucial’ ” to “the system of trial by jury.” Marshall v. Lonberger, 459 U.S. 422, 438, n. 6 (1983) (quoting Parker v. Randolph, 442 U.S. 62, 73 (1979)); accord, Francis v. Franklin, 471 U.S. 307, 324–325, n. 9 (1985).

A brief survey of prior decisions shows how firmly this Court has adhered to that practice. In Harris v. New York, 401 U.S. 222 (1971), the Court held that statements obtained from a defendant in violation of Miranda v. Arizona, 384 U.S. 436 (1966), could be introduced to impeach that defendant’s credibility, so long as the jury was instructed not to consider them as evidence of his guilt. In Walder v. United States, 347 U.S. 62 (1954), the Court affirmed the use of evidence obtained in violation of the Fourth Amendment for impeachment when the trial court had “carefully charged the jury” that it could not be considered evidence of guilt. Id., at 64. In Spencer v. Texas, 385 U.S. 554 (1967), the Court upheld the admission of evidence of the defendant’s prior criminal convictions for the purpose of sentence enhancement, provided that the jury was instructed that this evidence could not be used in determining guilt. In Watkins v. Sowders, 449 U.S. 341 (1981), the Court presumed that a jury could properly evaluate an eyewitness identification “under the instructions of the trial judge.” Id., at 347. And in Tennessee v. Street, 471 U.S. 409 (1985), the Court approved the admission of an accomplice’s incriminating confession given the “pointe[d] instruct[ions] [of] the trial court ‘not to consider the truthfulness of [the confession] in any way whatsoever.’ ” Id., at 414–415.

Most recently in Samia, we held that a limiting instruction was sufficient to defeat a Confrontation Clause claim. In that homicide case, evidence showed that Samia had traveled with his codefendant Stillwell to the Philippines to commit a murder for hire. Samia, 599 U. S., at 640. The trial court admitted Stillwell’s confession, which, as redacted, stated that he was in a van with some “other person ” when that person shot the victim, but the court told the jury that the confession could be considered only for the purpose of determining whether Stillwell himself was guilty. Id., at 642. Samia argued that admitting the confession even with the limiting instruction would inevitably prejudice him because “other evidence and statements at trial enabled the jury to immediately infer that the ‘other person’ described in the confession was Samia himself.”  Ibid. Nevertheless, we presumed that the jury was able to follow the limiting instruction, and we therefore affirmed Samia’s murder conviction.

Our cases have recognized only one situation in which a limiting instruction is insufficient: where a defendant is directly incriminated by the extrajudicial statements of a non-testifying codefendant. Bruton v. United States, 391 U.S. 123 (1968). We have declined to extend that exception, see Samia, 599 U. S., at 654, and the evidence in question in Bruton cases is worlds away from an expert’s basis testimony. If the Court thinks otherwise, it needs to explain why basis testimony falls into the Bruton category and creates a greater risk of juror confusion than all the other situations in which the Court has assumed that jurors are capable of following limiting instructions.


The Court’s assault on modern evidence law is not only wrongheaded; it is totally unnecessary. Today’s decision vacates the Arizona court’s judgment because the testifying expert’s testimony was hearsay. I agree with that bottom line, but not because of the majority’s novel theory that basis testimony is always hearsay. Rather, I would vacate and remand because the expert’s testimony is hearsay under any mainstream conception, including that of the Federal Rules of Evidence.

To understand why, begin with the facts. A state forensic scientist, Elizabeth Rast, tested items seized from the defendant and concluded that they were marijuana and methamphetamine. Rast took notes of her tests, see App. to Pet. for Cert. 88a–126a, and she signed a report confirming the results, see id., at 85a–87a. At trial, Rast was unavailable, so the prosecution called another forensic scientist, Greggory Longoni, to provide his expert opinion about the testing, and Longoni relied on Rast’s report in doing so.

Under Rules 703 and 705, Longoni could have offered his expert opinion that, based on the information in Rast’s report and notes, the items she tested contained marijuana or methamphetamine. In so answering, he would acknowledge that he relied on Rast’s report and lab notes to reach his opinion. He could have also disclosed the information in the report, if the court found that the probative value of that information substantially outweighed the risk of prejudice. See Fed. Rule Evid. 703. But he could not testify that any of the information in the report was correct—for instance, that Rast actually performed the tests she recorded or that she did so correctly. Nor could he testify that the items she tested were the ones seized from Smith. Longoni did not have personal knowledge of any of these facts, and it is unclear what “reliable” scientific “methods” could lead him to intuit their truth from Rast’s records. Fed. Rule Evid. 702(c) (defining a permissible expert opinion).

The strictures of the Federal Rules here track the requirements of our Confrontation Clause precedents. If Longoni testified to the truth of the fact that Rast actually performed the tests indicated in her report and notes and that she carried out those tests properly, he violated the Confrontation Clause—assuming, of course, that the notes were “testimonial,” a question that the Court does not reach. But he would also violate the Federal Rules, which do not allow experts to testify to the truth of inadmissible hearsay. In other words, except for the question whether Rast’s report was “testimonial,” the Federal Rules and the requirements of the Confrontation Clause are the same. This case thus offers no occasion to blow up the Federal Rules.

As it happens, I agree with the Court that Longoni stepped over the line and at times testified to the truth of the matter asserted. The prosecution asked Longoni on several occasions to describe the tests that Rast performed or to swear to their accuracy, and Longoni played along. He stated as fact that Rast followed the lab’s “typical intake process” and that she complied with the “policies and practices” of the lab. App. to Pet. for Cert. 40a–42a. He also testified that Rast used certain “scientific method[s]” to analyze the samples, such as performing certain tests or running a “blank.” Id., at 41a–42a, 46a–48a. By asserting these facts as true, Longoni effectively entered inadmissible hearsay into the record, thus implicating the Confrontation Clause. The Court could have said that—and stopped there.

*  *  *

For more than a half-century, the Federal Rules of Evidence and similar state rules have reasonably allowed experts to disclose the information underlying their opinion. Because the Court places this form of testimony in constitutional doubt in many cases, I concur only in the judgment.

1  I refer to the Federal Rules to illustrate the consequences of the Court’s opinion. The witness in this case testified in an Arizona state court, and his testimony was therefore governed by the relevant state rules, which are virtually identical to the Federal Rules. Of course, the Arizona courts are free to interpret those rules as they see fit, and I do not address the question whether the witness’s testimony was proper under Arizona law.

Jun 21, 2024 Judgment VACATED and case REMANDED. Kagan, J., delivered the opinion of the Court, in which Sotomayor, Kavanaugh, Barrett, and Jackson, JJ., joined, and in which Thomas and Gorsuch, JJ., joined as to Parts I, II, and IV. Thomas, J., and Gorsuch, J., filed opinions concurring in part. Alito, J., filed an opinion concurring in the judgment, in which Roberts, C. J., joined.
Jan 10, 2024 Argued. For petitioner: Hari Santhanam, Chicago, Ill. For United States, as amicus curiae: Eric J. Feigin, Deputy Solicitor General, Department of Justice, Washington, D. C. For respondent: Alexander W. Samuels, Principal Deputy Solicitor General, Phoenix, Ariz.
Dec 29, 2023 Reply of Jason Smith submitted.
Dec 29, 2023 Reply of petitioner Jason Smith filed. (Distributed)
Dec 20, 2023 Brief amici curiae of American Board of Forensic Toxicology, et al. filed. (Distributed)
Dec 20, 2023 Amicus brief of Oklahoma, Colorado, 36 Additional States, 1 Territory, and 1 Federal District submitted.
Dec 20, 2023 Amicus brief of Criminal Justice Legal Foundation submitted.
Dec 20, 2023 Amicus brief of National District Attorneys Association et al submitted.
Dec 20, 2023 Brief amici curiae of National District Attorneys Association et al filed. (Distributed)
Dec 20, 2023 Amicus brief of American Board of Forensic Toxicology, American Society of Crime Laboratory Directors, Association of Forensic Quality Assurance Managers, Center for Forensic Science Research and Education, International Association of Coroners and Medical Examiners, National Association of Medical Examiners, and Society of Forensic Toxicologists submitted.
Dec 20, 2023 Brief amici curiae of National District Attorneys Association, et al. filed. (Distributed)
Dec 20, 2023 Brief amici curiae of Colorado, et al. filed. (Distributed). (Corrected paper form of brief received and distributed--December 28, 2023--original electronic submission contains correction.)
Dec 20, 2023 Brief amici curiae of Colorado, et al. filed. (Distributed)
Dec 20, 2023 Brief amicus curiae of Criminal Justice Legal Foundation filed. (Distributed)
Dec 20, 2023 Brief amici curiae of Oklahoma, et al. filed. (Distributed)
Dec 13, 2023 Brief of Arizona submitted.
Dec 13, 2023 Respondent's Request to Lodge Non-Record Material Pursuant to Rule 32.3 of Arizona submitted.
Dec 13, 2023 Brief of respondent Arizona filed. (Distributed)
Dec 13, 2023 Respondent's request to lodge non-record material pursuant to Rule 32.3 filed.
Dec 11, 2023 Motion of the Solicitor General for leave to participate in oral argument as amicus curiae and for divided argument GRANTED.
Dec 7, 2023 Record received from the Court of Appeals of Arizona, Division One (one envelope). The remainder of the record is electronic and is available with the Clerk.
Nov 30, 2023 CIRCULATED
Nov 28, 2023 Motion of United States for leave to participate in oral argument and for divided argument submitted.
Nov 28, 2023 Motion of the Solicitor General for leave to participate in oral argument as amicus curiae and for divided argument filed.
Nov 28, 2023 Motion of United States for leave to participate in oral argument and for divided argument submitted.
Nov 21, 2023 Record requested from the Court of Appeals of Arizona, Division One.
Nov 20, 2023 Amicus brief of National Association of Criminal Defense Lawyers, American Civil Liberties Union, and American Civil Liberties Union Foundation of Arizona submitted.
Nov 20, 2023 Amicus brief of Alameda County Public Defender and California Public Defenders Association submitted.
Nov 20, 2023 Amicus brief of Richard D. Friedman submitted.
Nov 20, 2023 Brief amicus curiae of United States in support of neither party filed.
Nov 20, 2023 Brief amici curiae of The Innocence Network, Forensic Justice Project, et al.. filed.
Nov 20, 2023 Brief amicus curiae of Richard D. Friedman filed.
Nov 20, 2023 Brief amici curiae of Alameda County Public Defender, et al. filed.
Nov 20, 2023 Brief amici curiae of National Association of Criminal Defense Lawyers, et al. filed.
Nov 17, 2023 SET FOR ARGUMENT on Wednesday, January 10, 2024.
Nov 17, 2023 Amicus brief of National College for DUI Defense, Inc. submitted.
Nov 17, 2023 Brief amicus curiae of National College for DUI Defense, Inc. filed.
Nov 13, 2023 Joint Appendix submitted.
Nov 13, 2023 Brief of petitioner Jason Smith filed.
Nov 13, 2023 Joint appendix filed. (Statement of costs filed)
Nov 13, 2023 Brief of Jason Smith submitted.
Sep 29, 2023 Petition GRANTED.
Jul 12, 2023 DISTRIBUTED for Conference of 9/26/2023.
Jul 11, 2023 Reply of petitioner Jason Smith filed. (Distributed)
Jun 26, 2023 Brief of respondent Arizona in opposition filed.
May 19, 2023 Motion to extend the time to file a response is granted and the time is extended to and including June 26, 2023.
May 18, 2023 Motion to extend the time to file a response from May 25, 2023 to June 26, 2023, submitted to The Clerk.
Apr 25, 2023 Response Requested. (Due May 25, 2023)
Apr 19, 2023 DISTRIBUTED for Conference of 5/11/2023.
Apr 7, 2023 Waiver of right of respondent Arizona to respond filed.
Mar 14, 2023 Petition for a writ of certiorari filed. (Response due April 17, 2023)