Chief Justice Rehnquist
delivered the opinion of the Court.
In Chapman v. California, 386 U. S. 18, 24 (1967), we held that the standard for determining whether a conviction must be set aside because of federal constitutional error is whether the error “was harmless beyond a reasonable doubt.” In this case we must decide whether the Chapman harmless-error standard applies in determining whether the [623] prosecution’s use for impeachment purposes of petitioner’s post-Miranda1 silence, in violation of due process under Doyle v. Ohio, 426 U. S. 610 (1976), entitles petitioner to ha-beas corpus relief. We hold that it does not. Instead, the standard for determining whether habeas relief must be granted is whether the Doyle error “had substantial and injurious effect or influence in determining the jury’s verdict.” Kotteakos v. United States, 328 U. S. 750, 776 (1946). The Kotteakos harmless-error standard is better tailored to the nature and purpose of collateral review than the Chapman standard, and application of a less onerous harmless-error standard on habeas promotes the considerations underlying our habeas jurisprudence. Applying this standard, we conclude that petitioner is not entitled to habeas relief.
Petitioner Todd A. Brecht was serving time in a Georgia prison for felony theft when his sister and her husband, Molly and Roger Hartman, paid the restitution for petitioner’s crime and assumed temporary custody of him. The Hartmans brought petitioner home with them to Alma, Wisconsin, where he was to reside with them before entering a halfway house. This caused some tension in the Hartman household because Roger Hartman, a local district attorney, disapproved of petitioner’s heavy drinking habits and homosexual orientation, not to mention his previous criminal exploits. To make the best of the situation, though, the Hartmans told petitioner, on more than one occasion, that he was not to drink alcohol or engage in homosexual activities in their home. Just one week after his arrival, however, petitioner violated this house rule.
While the Hartmans were away, petitioner broke into their liquor cabinet and began drinking. He then found a rifle in an upstairs room and began shooting cans in the backyard. When Roger Hartman returned home from work, petitioner shot him in the back and sped off in Mrs. Hartman’s car. [624] Hartman crawled to a neighbor’s house to summon help. (The downstairs phone in the Hartmans’ house was inoperable because petitioner had taken the receiver on the upstairs phone off the hook.) Help came, but Hartman’s wound proved fatal. Meanwhile, petitioner had driven Mrs. Hartman’s car into a ditch in a nearby town. When a police officer stopped to offer assistance, petitioner told him that his sister knew about his car mishap and had called a tow truck. Petitioner then hitched a ride to Winona, Minnesota, where he was stopped by police. At first he tried to conceal his identity, but he later identified himself and was arrested. When he was told that he was being held for the shooting, petitioner replied that “it was a big mistake” and asked to talk with “somebody that would understand [him].” App. 39, 78. Petitioner was returned to Wisconsin, and thereafter was given his Miranda warnings at an arraignment.
Then petitioner was charged with first-degree murder. At trial in the Circuit Court for Buffalo County, he took the stand and admitted shooting Hartman, but claimed it was an accident. According to petitioner, when he saw Hartman pulling into the driveway on the evening of the shooting, he ran to replace the gun in the upstairs room where he had found it. But as he was running toward the stairs in the downstairs hallway, he tripped, causing the rifle to discharge the fatal shot. After the shooting, Hartman disappeared, so petitioner drove off in Mrs. Hartman’s car to find him. Upon spotting Hartman at his neighbor’s door, however, petitioner panicked and drove away.
The State argued that petitioner’s account was belied by the fact that he had failed to get help for Hartman, fled the Hartmans’ home immediately after the shooting, and lied to the police officer who came upon him in the ditch about having called Mrs. Hartman. In addition, the State pointed out that petitioner had failed to mention anything about the shooting being an accident to the officer who found him in the ditch, the man who gave him a ride to Winona, or the [625] officers who eventually arrested him. Over the objections of defense counsel, the State also asked petitioner during cross-examination whether he had told anyone at any time before trial that the shooting was an accident, to which petitioner replied “no,” and made several references to petitioner’s pretrial silence during closing argument.2 Finally, the State offered extrinsic evidence tending to contradict petitioner’s story, including the path the bullet traveled through Mr. Hartman’s body (horizontal to slightly downward) and the location where the rifle was found after the shooting (outside), as well as evidence of motive (petitioner’s hostility toward Mr. Hartman because of his disapproval of petitioner’s sexual orientation).
The jury returned a guilty verdict, and petitioner was sentenced to life imprisonment. The Wisconsin Court of [626] Appeals set the conviction aside on the ground that the State’s references to petitioner’s post-Miranda silence, see n. 2, supra, violated due process under Doyle v. Ohio, 426 U. S. 610 (1976), and that this error was sufficiently “prejudicial” to require reversal. State v. Brecht, 138 Wis. 2d 158, 168-169, 405 N. W. 2d 718, 723 (1987). The Wisconsin Supreme Court reinstated the conviction. Although it agreed that the State’s use of petitioner’s post-Miranda silence was impermissible, the court determined that this error “‘was harmless beyond a reasonable doubt.’” State v. Brecht, 143 Wis. 2d 297, 317, 421 N. W. 2d 96, 104 (1988) (quoting Chapman v. California, 386 U. S. 18, 24 (1967)). In finding the Doyle violation harmless, the court noted that the State’s “improper references to Brecht’s silence were infrequent,” in that they “comprised less than two pages of a 900 page transcript, or a few minutes in a four day trial in which twenty-five witnesses testified,” and that the State’s evidence of guilt was compelling. 143 Wis. 2d, at 317, 421 N. W. 2d, at 104.
Petitioner then sought a writ of habeas corpus under 28 U. S. C. § 2254, reasserting his Doyle claim. The District Court agreed that the State’s use of petitioner’s post-Miranda silence violated Doyle, but disagreed with the Wisconsin Supreme Court that this error was harmless beyond a reasonable doubt, and set aside the conviction. 759 F. Supp. 500 (WD Wis. 1991). The District Court based its harmless-error determination on its view that the State’s evidence of guilt was not “overwhelming,” and that the State’s references to petitioner’s post-Miranda silence, though “not extensive,” were “crucial” because petitioner’s defense turned on his credibility. Id., at 508. The Court of Appeals for the Seventh Circuit reversed. It, too, concluded that the State’s references to petitioner’s post-Miranda silence violated Doyle, but it disagreed with both the standard that the District Court had applied in conducting its harmless-error [627] inquiry and the result it reached. 944 F. 2d 1363, 1368, 1375-1376 (1991).
The Court of Appeals held that the Chapman harmless-error standard does not apply in reviewing Doyle error on federal habeas. Instead, because of the “prophylactic” nature of the Doyle rule, 944 F. 2d, at 1370, as well as the costs attendant to reversing state convictions on collateral review, id., at 1373, the Court of Appeals held that the standard for determining whether petitioner was entitled to habeas relief was whether the Doyle violation “ ‘had substantial and injurious effect or influence in determining the jury’s verdict,’ ” 944 F. 2d, at 1376 (quoting Kotteakos v. United States, 328 U. S., at 776). Applying this standard, the Court of Appeals concluded that petitioner was not entitled to relief because, “given the many more, and entirely proper, references to [petitioner’s] silence preceding arraignment,” he could not contend with a “straight face” that the State’s use of his post-Miranda silence had a “substantial and injurious effect” on the jury’s verdict. 944 F. 2d, at 1376.
We granted certiorari to resolve a conflict between Courts of Appeals on the question whether the Chapman harmless-error standard applies on collateral review of Doyle violations, 604 U. S. 972 (1992),3 and now affirm.
We are the sixth court to pass on the question whether the State’s use for impeachment purposes of petitioner’s post-Miranda silence requires reversal of his murder conviction. Petitioner urges us to even the count, and decide matters in his favor once and for all. He argues that the Chapman harmless-error standard applies with equal force on collateral review of Doyle error. According to petitioner, the need to prevent state courts from relaxing their standards on direct review of Doyle claims, and the confusion which would ensue were we to adopt the Kotteakos harmless-error standard on [628] collateral review, require application of the Chapman standard here. Before considering these arguments, however, we must first characterize the nature of Doyle error itself.
In Doyle v. Ohio, 426 U. S., at 619, we held that “the use for impeachment purposes of [a defendant’s] silence, at the time of arrest and after receiving Miranda warnings, violate[s] the Due Process Clause of the Fourteenth Amendment.” This rule “rests on ‘the fundamental unfairness of implicitly assuring a suspect that his silence will not be used against him and then using his silence to impeach an explanation subsequently offered at trial.’” Wainwright v. Greenfield, 474 U. S. 284, 291 (1986) (quoting South Dakota v. Neville, 459 U. S. 553, 565 (1983)). The “implicit assurance” upon which we have relied in our Doyle line of cases is the right-to-remain-silent component of Miranda. Thus, the Constitution does not prohibit the use for impeachment purposes of a defendant’s silence prior to arrest, Jenkins v. Anderson, 447 U. S. 231, 239 (1980), or after arrest if no Miranda warnings are given, Fletcher v. Weir, 455 U. S. 603, 606-607 (1982) (per curiam). Such silence is probative and does not rest on any implied assurance by law enforcement authorities that it will carry no penalty. See 447 U. S., at 239.
This case illustrates the point well. The first time petitioner claimed that the shooting was an accident was when he took the stand at trial. It was entirely proper — and probative — for the State to impeach his testimony by pointing out that petitioner had failed to tell anyone before the time he received his Miranda warnings at his arraignment about . the shooting being an accident. Indeed, if the shooting was an accident, petitioner had every reason — including to clear his name and preserve evidence supporting his version of the events — to offer his account immediately following the shooting. On the other hand, the State’s references to petitioner’s silence after that point in time, or more generally to petitioner’s failure to come forward with his version of [629] events at any time before trial, see n. 2, supra, crossed the Doyle line. For it is conceivable that, once petitioner had been given his Miranda warnings, he decided to stand on his right to remain silent because he believed his silence would not be used against him at trial.
The Court of Appeals characterized Doyle as “a prophylactic rule.” 944 F. 2d, at 1370. It reasoned that, since the need for Doyle stems from the implicit assurance that flows from Miranda warnings, and “the warnings required by Miranda are not themselves part of the Constitution,” “Doyle is ... a prophylactic rule designed to protect another prophylactic rule from erosion or misuse.” Ibid. But Doyle was not simply a further extension of the Miranda prophylactic rule. Rather, as we have discussed, it is rooted in fundamental fairness and due process concerns. However real these concerns, Doyle does not “ ‘overprotec[t]’ ” them. Duckworth v. Eagan, 492 U. S. 195, 209 (1989) (O’Connor, J., concurring). Under the rationale of Doyle, due process is violated whenever the prosecution uses for impeachment purposes a defendant’s post-Miranda silence. Doyle thus does not bear the hallmarks of a prophylactic rule.
Instead, we think Doyle error fits squarely into the category of constitutional violations which we have characterized as ‘“trial error.’” See Arizona v. Fulminante, 499 U. S. 279, 307 (1991). Trial error “occur[s] during the presentation of the case to the jury,” and is amenable to harmless-error analysis because it “may ... be quantitatively assessed in the context of other evidence presented in order to determine [the effect it had on the trial].” Id., at 307-308. At the other end of the spectrum of constitutional errors lie “structural defects in the constitution of the trial mechanism, which defy analysis by ‘harmless-error’ standards.” Id., at 309. The existence of such defects — deprivation of the right to counsel,4 for example — requires automatic reversal of the [630] conviction because they infect the entire trial process. See id., at 309-310. Since our landmark decision in Chapman v. California, 386 U. S. 18 (1967), we have applied the harmless-beyond-a-reasonable-doubt standard in reviewing claims of constitutional error of the trial type.
In Chapman, we considered whether the prosecution’s reference to the defendants’ failure to testify at trial, in violation of the Fifth Amendment privilege against self-incrimination,5 required reversal of their convictions. We rejected the argument that the Constitution requires a blanket rule of automatic reversal in the case of constitutional error, and concluded instead that “there may be some constitutional errors which in the setting of a particular case are so unimportant and insignificant that they may, consistent with the Federal Constitution, be deemed harmless.” Id., at 22. After examining existing harmless-error rules, including the federal rule (28 U. S. C. §2111), we held that “before a federal constitutional error can be held harmless, the court must be able to declare a belief that it was harmless beyond a reasonable doubt.” 386 U. S., at 24. The State bears the burden of proving that an error passes muster under this standard.
Chapman reached this Court on direct review, as have most of the cases in which we have applied its harmless-error standard. Although we have applied the Chapman standard in a handful of federal habeas cases, see, e. g., Yates v. Evatt, 500 U. S. 391 (1991); Rose v. Clark, 478 U. S. 570 (1986); Milton v. Wainwright, 407 U. S. 371 (1972); Anderson v. Nelson, 390 U. S. 523 (1968) (per curiam), we have yet squarely to address its applicability on collateral review.6 [631] Petitioner contends that we are bound by these habeas cases, by way of stare decisis, from holding that the Kotteakos harmless-error standard applies on habeas review of Doyle error. But since we have never squarely addressed the issue, and have at most assumed the applicability of the Chapman standard on habeas, we are free to address the issue on the merits. See Edelman v. Jordan, 415 U. S. 651, 670-671 (1974).
The federal habeas corpus statute is silent on this point. It permits federal courts to entertain a habeas petition on behalf of a state prisoner “only on the ground that he is in custody in violation of the Constitution or laws or treaties of the United States,” 28 U. S. C. § 2254(a), and directs simply that the court “dispose of the matter as law and justice require,” § 2243. The statute says nothing about the standard for harmless-error review in habeas cases. Respondent urges us to fill this gap with the Kotteakos standard, under which an error requires reversal only if it “had substantial and injurious effect or influence in determining the jury’s verdict.” Kotteakos v. United States, 328 U. S., at 776. This standard is grounded in the federal harmless-error statute. 28 U. S. C. §2111 (“On the hearing of any appeal or writ of certiorari in any case, the court shall give judgment after an examination of the record without regard to errors or defects which do not affect the substantial rights of the parties”).7 On its face § 2111 might seem to address the situ[632] ation at hand, but to date we have limited its application to claims of nonconstitutional error in federal criminal cases. See, e. g., United States v. Lane, 474 U. S. 438 (1986).
Petitioner asserts that Congress’ failure to enact various proposals since Chapman was decided that would have limited the availability of habeas relief amounts to legislative disapproval of application of a less stringent harmless-error standard on collateral review of constitutional error. Only one of these proposals merits discussion here. In 1972, a bill was proposed that would have amended 28 U. S. C. § 2264 to require habeas petitioners to show that “ ‘a different result would probably have obtained if such constitutional violation had not occurred.’” 118 Cong. Rec. 24936 (1972) (quoting S. 3833, 92d Cong., 2d Sess. (1972)). In response, the Attorney General suggested that the above provision be modified to make habeas relief available only where the petitioner “‘suffered a substantial deprivation of his constitutional rights at his trial.’” 118 Cong. Rec. 24939 (1972) (quoting letter from Richard G. Kleindienst, Attorney General, to Emanuel Celler, Chairman of the House Committee on the Judiciary (June 21,1972)). This language of course parallels the federal harmless-error rule. But neither the Attorney General’s suggestion nor the proposed bill itself was ever enacted into law.
As a general matter, we are “reluctant to draw inferences from Congress’ failure to act.” Schneidewind v. ANR Pipeline Co., 485 U. S. 293, 306 (1988) (citing American Trucking Assns., Inc. v. Atchison, T. & S. F. R. Co., 387 U. S. 397, [633]*633416-418 (1967)); Red Lion Broadcasting Co. v. FCC, 396 U. S. 367, 381, n. 11 (1969)). We find no reason to depart from this rule here. In the absence of any express statutory guidance from Congress, it remains for this Court to determine what harmless-error standard applies on collateral review of petitioner’s Doyle claim. We have filled the gaps of the habeas corpus statute with respect to other matters, see, e. g., McCleskey v. Zant, 499 U. S. 467, 487 (1991); Wainwright v. Sykes, 433 U. S. 72, 81 (1977); Sanders v. United States, 373 U. S. 1, 15 (1963); Townsend v. Sain, 372 U. S. 293, 312-313 (1963), and find it necessary to do so here. As always, in defining the scope of the writ, we look first to the considerations underlying our habeas jurisprudence, and then determine whether the proposed rule would advance or inhibit these considerations by weighing the marginal costs and benefits of its application on collateral review.
The principle that collateral review is different from direct review resounds throughout our habeas jurisprudence. See, e. g., Wright v. West, 505 U. S. 277, 292-293 (1992) (opinion of Thomas, J.); Teague v. Lane, 489 U. S. 288, 306 (1989) (opinion of O’Connor, J.); Pennsylvania v. Finley, 481 U. S. 551, 556-557 (1987); Mackey v. United States, 401 U. S. 667, 682 (1971) (Harlan, J., concurring in judgments in part and dissenting in part). Direct review is the principal avenue for challenging a conviction. “When the process of direct review— which, if a federal question is involved, includes the right to petition this Court for a writ of certiorari — comes to an end, a presumption of finality and legality attaches to the conviction and sentence. The role of federal habeas proceedings, while important in assuring that constitutional rights are observed, is secondary and limited. Federal courts are not forums in which to relitigate state trials.” Barefoot v. Estelle, 463 U. S. 880,. 887 (1983).
In keeping with this distinction, the writ of habeas corpus has historically been regarded as an extraordinary remedy, “a bulwark against convictions that violate 'fundamental fair[634] ness.’” Engle v. Isaac, 456 U. S. 107, 126 (1982) (quoting Wainwright v. Sykes, supra, at 97 (Stevens, J., concurring)). “Those few who are ultimately successful [in obtaining habeas relief] are persons whom society has grievously wronged and for whom belated liberation is little enough compensation.” Fay v. Noia, 372 U. S. 391, 440-441 (1963). See also Kuhlmann v. Wilson, 477 U. S. 436, 447 (1986) (plurality opinion) (“The Court uniformly has been guided by the proposition that the writ should be available to afford relief to those ‘persons whom society has grievously wronged’ in light of modern concepts of justice”) (quoting Fay v. Noia, supra, at 440-441); Jackson v. Virginia, 443 U. S. 307, 332, n. 5 (1979) (Stevens, J., concurring in judgment) (Habeas corpus “is designed to guard against extreme malfunctions in the state criminal justice systems”). Accordingly, it hardly bears repeating that “ ‘an error that may justify reversal on direct appeal will not necessarily support a collateral attack on a final judgment.’ ” United States v. Frady, 456 U. S. 152, 165 (1982) (quoting United States v. Addonizio, 442 U. S. 178, 184 (1979)).8
Recognizing the distinction between direct and collateral review, we have applied different standards on habeas than would be applied on direct review with respect to matters other than harmless-error analysis. Our recent retroactivity jurisprudence is a prime example. Although new rules always have retroactive application to criminal cases on direct review, Griffith v. Kentucky, 479 U. S. 314, 320-328 (1987), we have held that they seldom have retroactive application to criminal cases on federal habeas, Teague v. Lane, supra, at 305-310 (opinion of O’Connor, J.). Other examples abound throughout our habeas cases. See, e.g., Pennsylvania v.
[635] Finley, 481 U. S. 551, 555-556 (1987) (Although the Constitution guarantees the right to counsel on direct appeal, Douglas v. California, 372 U. S. 353, 355 (1963), there is no “right to counsel when mounting collateral attacks”); United States v. Frady, supra, at 162-169 (While the federal “plain error” rule applies in determining whether a defendant may raise a claim for the first time on direct appeal, the “cause and prejudice” standard applies in determining whether that same claim may be raised on habeas); Stone v. Powell, 428 U. S. 465, 489-496 (1976) (Claims under Mapp v. Ohio, 367 U. S. 643 (1961), are not cognizable on habeas as long as the state courts have provided a full and fair opportunity to litigate them at trial or on direct review).
The reason most frequently advanced in our cases for distinguishing between direct and collateral review is the State’s interest in the finality of convictions that have survived direct review within the state court system. See, e. g., Wright v. West, supra, at 293 (opinion of Thomas, J.); McCleskey v. Zant, 499 U. S., at 491; Wainwright v. Sykes, 433 U. S., at 90. We have also spoken of comity and federalism. “The States possess primary authority for defining and enforcing the criminal law. In criminal trials they also hold the initial responsibility for vindicating constitutional rights. Federal intrusions into state criminal trials frustrate both the States’ sovereign power to punish offenders and their good-faith attempts to honor constitutional rights.” Engle v. Isaac, supra, at 128. See also Coleman v. Thompson, 501 U. S. 722, 748 (1991); McCleskey, supra, at 491. Finally, we have recognized that “[Iliberal allowance of the writ.. . degrades the prominence of the trial itself,” Engle, supra, at 127, and at the same time encourages habeas petitioners to relitigate their claims on collateral review, see Rose v. Lundy, 455 U. S. 509, 547 (1982) (Stevens, J., dissenting).
In light of these considerations, we must decide whether the same harmless-error standard that the state courts ap[636] plied on direct review of petitioner’s Doyle claim also applies in this habeas, proceeding. We are the sixth court to pass on the question whether the State’s use for impeachment purposes of petitioner’s post-Miranda silence in this case requires reversal of his conviction. Each court that has reviewed the record has disagreed with the court before it as to whether the State’s Doyle error was “harmless.” State courts are fully qualified to identify constitutional error and evaluate its prejudicial effect on the trial process under Chapman, and state courts often occupy a superior vantage point from which to evaluate the effect of trial error. See Rushen v. Spain, 464 U. S. 114, 120 (1983) (per curiam). For these reasons, it scarcely seems logical to require federal habeas courts to engage in the identical approach to harmless-error review that Chapman requires state courts to engage in on direct review.
Petitioner argues that application of the Chapman harmless-error standard on collateral review is necessary to deter state courts from relaxing their own guard in reviewing constitutional error and to discourage prosecutors from committing error in the first place. Absent affirmative evidence that state-court judges are ignoring their oath, we discount petitioner’s argument that courts will respond to our ruling by violating their Article VI duty to uphold the Constitution. See Robb v. Connolly, 111 U. S. 624, 637 (1884). Federalism, comity, and the constitutional obligation of state and federal courts all counsel against any presumption that a decision of this Court will “deter” lower federal or state courts from fully performing their sworn duty. See Engle, supra, at 128; Schneckloth v. Bustamonte, 412 U. S. 218, 263-265 (1973) (Powell, J., concurring). In any event, we think the costs of applying the Chapman standard on federal habeas outweigh the additional deterrent effect, if any, that would be derived from its application on collateral review.
[637] Overturning final and presumptively correct convictions on collateral review because the State cannot prove that an error is harmless under Chapman undermines the States’ interest in finality and infringes upon their sovereignty over criminal matters. Moreover, granting habeas relief merely because there is a “ ‘reasonable possibility’ ” that trial error contributed to the verdict, see Chapman v. California, 386 U. S., at 24 (quoting Fahy v. Connecticut, 375 U. S. 85, 86 (1963)), is at odds with the historic meaning of habeas corpus — to afford relief to those whom society has “grievously wronged.” Retrying defendants whose convictions are set aside also imposes significant “social costs,” including the expenditure of additional time and resources for all the parties involved, the “erosion of memory” and “dispersion of witnesses” that accompany the passage of time and make obtaining convictions on retrial more difficult, and the frustration of “society’s interest in the prompt administration of justice.” United States v. Mechanik, 475 U. S. 66, 72 (1986) (internal quotation marks omitted). And since there is no statute of limitations governing federal habeas, and the only laches recognized is that which affects the State’s ability to defend against the claims raised on habeas, retrials following the grant of habeas relief ordinarily take place much later than do retrials following reversal on direct review.
The imbalance of the costs and benefits of applying the Chapman harmless-error standard on collateral review counsels in favor of applying a less onerous standard on habeas review of constitutional error. The Kotteakos standard, we believe, fills the bill. The test under Kotteakos is whether the error “had substantial and injurious effect or influence in determining the jury’s verdict.” 328 U. S., at 776. Under this standard, habeas petitioners may obtain plenary review of their constitutional claims, but they are not entitled to habeas relief based on trial error unless they can establish that it resulted in “actual prejudice.” See United States v. Lane, 474 U. S. 438, 449 (1986). The Kotteakos [638] standard is thus better tailored to the nature and purpose of collateral review and more likely to promote the considerations underlying our recent habeas cases. Moreover, because the Kotteakos standard is grounded in the federal harmless-error rule, 28 U. S. C. §2111, federal courts may turn to an existing body of case law in applying it. Therefore, contrary to the assertion of petitioner, application of the Kotteakos standard on collateral review is unlikely to confuse matters for habeas courts.
For the foregoing reasons, then, we hold that the Kot-teakos harmless-error standard applies in determining whether habeas relief must be granted because of constitutional error of the trial type.9 All that remains to be decided is whether petitioner is entitled to relief under this standard based on the State’s Doyle error. Because the Court of Appeals applied the Kotteakos standard below, we proceed to this question ourselves rather than remand the case for a new harmless-error determination. Cf. Yates v. Evatt, 500 U. S. 391, 407 (1991). At trial, petitioner admitted shooting Hartman, but claimed it was an accident. The principal question before the jury, therefore, was whether the State met its burden in proving beyond a reasonable doubt that the shooting was intentional. Our inquiry here is whether, in light of the record as a whole, the State’s improper use for impeachment purposes of petitioner’s post-Miranda silence, see n. 2, supra, “had substantial and injurious effect or influence in determining the jury’s verdict.” We think it clear that it did not.
[639] The State’s references to petitioner’s post-Miranda silence were infrequent, comprising less than two pages of the 900-page trial transcript in this case. And in view of the State’s extensive and permissible references to petitioner’s pre-Miranda silence — i e., his failure to mention anything about the shooting being an accident to the officer who found him in the ditch, the man who gave him a ride to Winona, or the officers who eventually arrested him — its references to petitioner’s post-Miranda silence were, in effect, cumulative. Moreover, the State’s evidence of guilt was, if not overwhelming, certainly weighty. The path of the bullet through Mr. Hartman’s body was inconsistent with petitioner’s testimony that the rifle had discharged as he was falling. The police officers who searched the Hartmans’ home found nothing in the downstairs hallway that could have caused petitioner to trip. The rifle was found outside the house (where Hartman was shot), not inside where petitioner claimed it had accidently fired, and there was a live round rammed in the gun’s chamber, suggesting that petitioner had tried to fire a second shot. Finally, other circumstantial evidence, including the motive proffered by the State, also pointed to petitioner’s guilt.
In light of the foregoing, we conclude that the Doyle error that occurred at petitioner’s trial did not “substantially]... influence” the jury’s verdict. Petitioner is therefore not entitled to habeas relief, and the judgment of the Court of Appeals is
Affirmed.