sure unless his omission is of sufficient significance to result in the denial of the defendant's right to a fair trial." 427 U. S., at 108. In Brady and Agurs, the prosecutor failed to disclose exculpatory evidence. In the present case, the prosecutor failed to disclose evidence that the defense might have used to impeach the Government's witnesses by showing bias or interest. Impeachment evidence, however, as well as exculpatory evidence, falls within the Brady rule. See Giglio v. United States, 405 U. S. 150, 154 (1972). Such evidence is "evidence favorable to an accused," Brady, 373 U. S., at 87, so that, if disclosed and used effectively, it may make the difference between conviction and acquittal. Cf. Napue v. Illinois, 360 U. S. 264, 269 (1959) ("The jury's estimate of the truthfulness and reliability of a given witness may well be determinative of guilt or innocence, and it is upon such subtle factors as the possible interest of the witness in testifying falsely that a defendant's life or liberty may depend"). The Court of Appeals treated impeachment evidence as constitutionally different from exculpatory evidence. According to that court, failure to disclose impeachment evidence is "even more egregious" than failure to disclose exculpatory evidence "because it threatens the defendant's right to confront adverse witnesses." 719 F. 2d, at 1464. Relying on Davis v. Alaska, 415 U. S. 308 (1974), the Court of Appeals held that the Government's failure to disclose requested impeachment evidence that the defense could use to conduct an effective cross-examination of important prosecution witnesses constitutues "constitutional error of the first magnitude'" requiring automatic reversal. 719 F. 2d, at 1464 (quoting Davis v. Alaska, supra, at 318). This Court has rejected any such distinction between impeachment evidence and exculpatory evidence. In Giglio v. United States, supra, the Government failed to disclose impeachment evidence similar to the evidence at issue in the present case, that is, a promise made to the key Government witness that he would not be prosecuted if he testified for the Government. This Court said: "When the 'reliability of a given witness may well be 999 405 Thus, the Court of Appeals' holding is inconsistent with our precedents. Moreover, the court's reliance on Davis v. Alaska for its "automatic reversal" rule is misplaced. In Davis, the defense sought to cross-examine a crucial prosecution witness concerning his probationary status as a juvenile delinquent. The defense intended by this cross-examination to show that the witness might have made a faulty identification of the defendant in order to shift suspicion away from himself or because he feared that his probationary status would be jeopardized if he did not satisfactorily assist the police and prosecutor in obtaining a conviction. Pursuant to a state rule of procedure and a state statute making juvenile adjudications inadmissible, the trial judge prohibited the defense from conducting the cross-examination. This Court reversed the defendant's conviction, ruling that the direct restriction on the scope of cross-examination denied the defendant "the right of effective cross-examination which ""would be constitutional error of the first magnitude and no amount of showing of want of prejudice would cure it." Brookhart v. Janis, 384 U. S. 1, 3."" 415 U. S., at 318 (quoting Smith Opinion of BLACKMUN, J. v. Illinois, 390 U. S. 129, 131 (1968)). States v. Cronic, 466 U. S. 648, 659 (1984). 473 U. S. See also United The present case, in contrast, does not involve any direct restriction on the scope of cross-examination. The defense was free to cross-examine the witnesses on any relevant subject, including possible bias or interest resulting from inducements made by the Government. The constitutional error, if any, in this case was the Government's failure to assist the defense by disclosing information that might have been helpful in conducting the cross-examination. As discussed above, such suppression of evidence amounts to a constitutional violation only if it deprives the defendant of a fair trial. Consistent with "our overriding concern with the justice of the finding of guilt," United States v. Agurs, 427 U. S., at 112, a constitutional error occurs, and the conviction must be reversed, only if the evidence is material in the sense that its suppression undermines confidence in the outcome of the trial. III It remains to determine the standard of materiality applicable to the nondisclosed evidence at issue in this case. Our starting point is the framework for evaluating the materiality of Brady evidence established in United States v. Agurs. The Court in Agurs distinguished three situations involving the discovery, after trial, of information favorable to the accused that had been known to the prosecution but unknown to the defense. The first situation was the prosecutor's knowing use of perjured testimony or, equivalently, the prosecutor's knowing failure to disclose that testimony used to convict the defendant was false. The Court noted the wellestablished rule that "a conviction obtained by the knowing use of perjured testimony is fundamentally unfair, and must be set aside if there is any reasonable likelihood that the false testimony could have affected the judgment of the jury." 667 Opinion of BLACKMUN, J. 427 U. S., at 103 (footnote omitted). Although this rule is stated in terms that treat the knowing use of perjured testimony as error subject to harmless-error review, it may as In fact, the Brady rule has its roots in a series of cases dealing with convictions based on the prosecution's knowing use of perjured testimony. In Mooney v. Holohan, 294 U. S. 103 (1935), the Court established the rule that the knowing use by a state prosecutor of perjured testimony to obtain a conviction and the deliberate suppression of evidence that would have impeached and refuted the testimony constitutes a denial of due process. The Court reasoned that "a deliberate deception of court and jury by the presentation of testimony known to be perjured" is inconsistent with "the rudimentary demands of justice." Id., at 112. The Court reaffirmed this principle in broader terms in Pyle v. Kansas, 317 U. S. 213 (1942), where it held that allegations that the prosecutor had deliberately suppressed evidence favorable to the accused and had knowingly used perjured testimony were sufficient to charge a due process violation. The Court again reaffirmed this principle in Napue v. Illinois, 360 U. S. 264 (1959). In Napue, the principal witness for the prosecution falsely testified that he had been promised no consideration for his testimony. The Court held that the knowing use of false testimony to obtain a conviction violates due process regardless of whether the prosecutor solicited the false testimony or merely allowed it to go uncorrected when it appeared. The Court explained that the principle that a State may not knowingly use false testimony to obtain a conviction-even false testimony that goes only to the credibility of the witness-is "implicit in any concept of ordered liberty." Id., at 269. Finally, the Court held that it was not bound by the state court's determination that the false testimony "could not in any reasonable likelihood have affected the judgment of the jury." Id., at 271. The Court conducted its own independent examination of the record and concluded that the false testimony "may have had an effect on the outcome of the trial." Id., at 272. Accordingly, the Court reversed the judgment of conviction. The rule that a conviction obtained by the knowing use of perjured testimony must be set aside if there is any reasonable likelihood that the false testimony could have affected the jury's verdict derives from Napue v. Illinois, 360 U. S., at 271. See n. 8, supra. See also Giglio v. United States, 405 U. S. 150, 154 (1972) (quoting Napue, 360 U. S., at 271). Napue antedated Chapman v. California, 386 U. S. 18 (1967), where the "harmless beyond a reasonable doubt" standard was established. The Court in Chapman noted that there was little, if any, difference between Opinion of BLACKMUN, J. 473 U. S. easily be stated as a materiality standard under which the fact that testimony is perjured is considered material unless failure to disclose it would be harmless beyond a reasonable doubt. The Court in Agurs justified this standard of materiality on the ground that the knowing use of perjured testimony involves prosecutorial misconduct and, more importantly, involves "a corruption of the truth-seeking function of the trial process." Id., at 104. At the other extreme is the situation in Agurs itself, where the defendant does not make a Brady request and the prosecutor fails to disclose certain evidence favorable to the accused. The Court rejected a harmless-error rule in that situation, because under that rule every nondisclosure is treated as error, thus imposing on the prosecutor a constitutional duty to deliver his entire file to defense counsel.10 427 U. S., at 111-112. At the same time, the Court rejected a standard that would require the defendant to demonstrate that the evidence if disclosed probably would have resulted in acquittal. Id., at 111. The Court reasoned: "If the standard applied to the usual motion for a new trial based on newly discovered evidence were the same when the evidence was in the State's possession as when it was found in a neutral source, there would be no special significance to the prosecutor's obligation to serve the cause of justice." Ibid. The a rule formulated, as in Napue, in terms of "whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction,'" and a rule "requiring the beneficiary of a constitutional error to prove beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained."" 386 U. S., at 24 (quoting Fahy v. Connecticut, 375 U. S. 85, 86-87 (1963)). It is therefore clear, as indeed the Government concedes, see Brief for United States 20, and 36-38, that this Court's precedents indicate that the standard of review applicable to the knowing use of perjured testimony is equivalent to the Chapman harmless-error standard. 10 This is true only if the nondisclosure is treated as error subject to harmless-error review, and not if the nondisclosure is treated as error only if the evidence is material under a not "harmless beyond a reasonable doubt" standard. |