delivered the opinion of the Court.

Respondent Robert Van Arsdall was convicted of murder in a Delaware trial court. The Supreme Court of Delaware reversed his conviction on the ground that the trial court, by improperly restricting defense counsel’s cross-examination designed to show bias on the part of a prosecution witness, had violated respondent’s confrontation rights under the Sixth and Fourteenth Amendments to the United States Constitution, and that such violation required automatic reversal. 486 A. 2d 1 (1984). While we agree that the trial court’s ruling was contrary to the mandate of the Confrontation Clause of the Sixth Amendment, we conclude that the Supreme Court of Delaware was wrong when it declined to consider whether that ruling was harmless in the context of the trial as a whole.

Shortly after midnight on January 1, 1982, Doris Epps was stabbed to death in an apartment in Smyrna, Delaware, after an all-day New Year’s Eve party. Respondent and Daniel Pregent, who by respondent’s testimony were the only two people in the apartment with Epps at the time she was killed, were arrested at the scene of the crime and charged with Epps’ murder. At separate trials, respondent was convicted and Pregent was acquitted.

The State’s case against respondent was based on circumstantial evidence, and proceeded on the theory that respondent had either killed Epps or assisted Pregent in doing so. Several of the partygoers testified about the party and the scene after the killing. The party, which lasted from late in the morning of December 31, 1981, until shortly before midnight, was held in the adjacent apartments of Pregent and *675Robert Fleetwood. Respondent, who was one of at least a dozen guests who attended the party during the course of the day, had stopped in for two brief periods in the late afternoon and early evening and then returned for a third time at about 11:30 p.m. By that time the party was over. Pregent had quarreled with a female guest, kicked a hole in a hallway wall and had to be restrained. An intoxicated Epps had been placed on a sofa bed in Pregent’s apartment after passing out. And shortly before 11 p.m., a second altercation of some kind occurred, prompting Fleetwood to close the party in his apartment to everyone except his two friends, Alice Meinier and Mark Mood. When respondent returned to Pregent’s apartment at about 11:30, only Pregent and Epps were present.

Robert Fleetwood was the 10th of 16 prosecution witnesses. In addition to recounting uncontroverted facts about the party, he testified that sometime between 11 and 11:30 p.m. he walked across the hall, looked into Pregent’s living room from the doorway, and saw respondent sitting on the edge of the sofa bed next to Pregent’s feet. Fleetwood, who did not have a complete view of the bed, did not see Epps or anyone else in the apartment. Upon returning to his own apartment, Fleetwood stayed awake long enough to hear nearby bells ring in the New Year, at which point he passed out. App. 82-85.

Meinier, who with Mood had remained awake in Fleet-wood’s apartment, testified that at roughly 1 a.m. respondent knocked at Fleetwood’s door. Respondent’s shirt and hands were spattered with blood, and he was holding a long, blood-covered knife. According to Meinier, respondent stated that “he had gotten in a fight” but that he “got them back.” Id., at 130. After turning the knife over to Mood and washing his hands, respondent said “I think there’s something wrong across the hall.” Id., at 132. Meinier went to Pregent’s apartment and discovered Epps’ body lying in a pool of blood on the kitchen floor. Mood then summoned the police, who found respondent in Fleetwood’s *676apartment and Pregent asleep on the blood-splattered sofa bed in his living room.

In addition to the testimony of the partygoers and the arresting officers, the State introduced Pregent’s postarrest statement, respondent’s two postarrest statements, and the testimony of a forensic expert. Among other things, the expert testified about the nature and source of the bloodstains on respondent’s clothing.

During Fleetwood’s cross-examination, defense counsel sought to impeach Fleetwood by questioning him about the dismissal of a criminal charge against him — being drunk on a highway — after he had agreed to speak with the prosecutor about Epps’ murder. When the prosecutor objected, the trial court allowed counsel to question Fleetwood on the matter outside the presence of the jury. Fleetwood acknowledged that the drunkenness charge had been dropped in exchange for his promise to speak with the prosecutor about the murder, but he denied that the agreement had affected his testimony.1 The trial court barred any cross-examination about that agreement, citing Delaware Rule of Evidence 403.2 The court also refused to permit defense counsel to cross-examine Fleetwood about his being questioned by the police in connection with an unrelated homicide that had occurred after Epps’ murder. On voir dire conducted outside the presence of the jury, Fleetwood denied that he had been *677offered any favors, inducements, promises, or deals concerning that homicide investigation in exchange for his testimony at respondent’s trial.

Respondent was the only defense witness. Consistent with his second statement to the police, he attributed Epps’ murder to Pregent. Consistent with Fleetwood’s testimony, he stated that he had returned to Pregent’s apartment, after drinking with friends, by about 11:30 p.m.

Defense counsel admitted in their opening and closing arguments to the jury that respondent was in Pregent’s apartment when Epps was killed. In closing argument, after attempting to discredit Fleetwood’s testimony (largely by emphasizing his intoxication), counsel stressed that all that testimony proved was what respondent “never denied,” that “he was at Danny Pregent’s apartment before Doris Epps was murdered.” App. 188-189. The jury found respondent guilty of first-degree murder and possession of a deadly weapon during the commission of a felony.

bn appeal, the Delaware Supreme Court reversed respondent’s conviction on the authority of the Confrontation Clause. Noting that “the bias of a witness is subject to exploration at trial and is ‘always relevant as discrediting the witness and affecting the weight of his testimony,”’ 486 A. 2d, at 6 (quoting Davis v. Alaska, 415 U. S. 308, 316 (1974)), the court found that the trial judge’s ruling denied respondent his constitutional right to effective cross-examination. By barring any cross-examination of Fleetwood about the dismissal of the public drunkenness charge, the ruling kept from the jury facts concerning bias that were central to assessing Fleetwood’s reliability. The court rejected the State’s argument that since “Fleetwood’s basic testimony was cumulative in nature and unimportant,” the Confrontation Clause error was harmless beyond a reasonable doubt. 486 A. 2d, at 7. The court held that “a blanket prohibition against exploring potential bias through cross-examination” *678is “a per se error,” so that “the actual prejudicial impact of such an error is not examined.” Ibid.3

We granted certiorari, 473 U. S. 923 (1985), and now vacate and remand.

The Confrontation Clause of the Sixth Amendment guarantees the right of an accused in a criminal prosecution “to be confronted with the witnesses against him.” The right of confrontation, which is secured for defendants in state as well as federal criminal proceedings, Pointer v. Texas, 380 U. S. 400 (1965), “means more than being allowed to confront the witness physically.” Davis v. Alaska, 415 U. S., at 315. Indeed, “‘[t]he main and essential purpose of confrontation is to secure for the opponent the opportunity of cross-examination.’” Id., at 315-316 (quoting 5 J. Wig-more, Evidence § 1395, p. 123 (3d ed. 1940)) (emphasis in original). Of particular relevance here, “[w]e have recognized that the exposure of a witness’ motivation in testifying is *679a proper and important function of the constitutionally protected right of cross-examination.” Davis, supra, at 316-317 (citing Greene v. McElroy, 360 U. S. 474, 496 (1959)). It does not follow, of course, that the Confrontation Clause of the Sixth Amendment prevents a trial judge from imposing any limits on defense counsel’s inquiry into the potential bias of a prosecution witness. On the contrary, trial judges retain wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on such cross-examination based on concerns about, among other things, harassment, prejudice, confusion of the issues, the witness’ safety, or interrogation that is repetitive or only marginally relevant. And as we observed earlier this Term, “the Confrontation Clause guarantees an opportunity for effective cross-examination, not cross-examination that is effective in whatever way, and to whatever extent, the defense might wish.” Delaware v. Fensterer, 474 U. S. 15, 20 (1985) (per curiam) (emphasis in original).

In this case, however, the trial court prohibited all inquiry into the possibility that Fleetwood would be biased as a result of the State’s dismissal of his pending public drunkenness charge. By thus cutting off all questioning about an event that the State conceded had taken place and that a jury might reasonably have found furnished the witness a motive for favoring the prosecution in his testimony, the court’s ruling violated respondent’s rights secured by the Confrontation Clause.4

The State somewhat tentatively suggests that a defendant should have to show “outcome determinative” prejudice in order to state a violation of the Confrontation Clause: Unless the particular limitation on cross-examination created a reasonable possibility that the jury returned an inaccurate guilty *680verdict, that limitation would not violate the Confrontation Clause. We disagree. While some constitutional claims by their nature require a showing of prejudice with respect to the trial as a whole, see, e. g., Strickland v. Washington, 466 U. S. 668 (1984) (ineffective assistance of counsel), the focus of the Confrontation Clause is on individual witnesses. Accordingly, the focus of the prejudice inquiry in determining whether the confrontation right has been violated must be on the particular witness, not on the outcome of the entire trial. It would be a contradiction in terms to conclude that a defendant denied any opportunity to cross-examine the witnesses against him nonetheless had been afforded his right to “confrontation]” because use of that right would not have affected the jury’s verdict. We think that a criminal defendant states a violation of the Confrontation Clause by showing that he was prohibited from engaging in otherwise appropriate cross-examination designed to show a prototypical form of bias on the part of the witness, and thereby “to expose to the jury the facts from which jurors . . . could appropriately draw inferences relating to the reliability of the witness.” Davis v. Alaska, supra, at 318. Respondent has met that burden here: A reasonable jury might have received a significantly different impression of Fleetwood’s credibility had respondent’s counsel been permitted to pursue his proposed line of cross-examination.

After concluding that the trial judge’s ruling was constitutional error, the Delaware Supreme Court rebuffed the State’s effort to show “beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained,” Chapman v. California, 386 U. S. 18, 24 (1967). In so doing, it offered no explanation why the Chapman harmless-error standard, which we have applied in other Confrontation Clause cases, e. g., Harrington v. California, 395 U. S. 250 (1969); Schneble v. Florida, 405 U. S. 427 (1972), is inapplicable here. We find respondent’s efforts to defend the automatic reversal rule unconvincing.

*681As we have stressed on more than one occasion, the Constitution entitles a criminal defendant to a fair trial, not a perfect one. E. g., United States v. Hasting, 461 U. S. 499, 508-509 (1983); Bruton v. United States, 391 U. S. 123, 135 (1968). In Chapman, this Court rejected the argument that all federal constitutional errors, regardless of their nature or the circumstances of the case, require reversal of a judgment of conviction. The Court reasoned that in the context of a particular case, certain constitutional errors, no less than other errors, may have been “harmless” in terms of their effect on the factfinding process at trial. Since Chapman, we have repeatedly reaffirmed the principle that an otherwise valid conviction should not be set aside if the reviewing court may confidently say, on the whole record, that the constitutional error was harmless beyond a reasonable doubt. E. g., United States v. Hasting, supra (improper comment on defendant’s silence at trial); Moore v. Illinois, 434 U. S. 220, 232 (1977) (admission of identification obtained in violation of right to counsel); Harrington v. California, supra (admission of nontestifying codefendant’s statement). The harmless-error doctrine recognizes the principle that the central purpose of a criminal trial is to decide the factual question of the defendant’s guilt or innocence, United States v. Nobles, 422 U. S. 225, 230 (1975), and promotes public respect for the criminal process by focusing on the underlying fairness of the trial rather than on the virtually inevitable presence of immaterial error. Cf. R. Traynor, The Riddle of Harmless Error 50 (1970) (“Reversal for error, regardless of its effect on the judgment, encourages litigants to abuse the judicial process and bestirs the public to ridicule it”).

At the same time, we have observed that some constitutional errors — such as denying a defendant the assistance of counsel at trial, or compelling him to stand trial before a trier of fact with a financial stake in the outcome — are so fundamental and pervasive that they require reversal without regard to the facts or circumstances of the particular case. *682Chapman, supra, at 23, n. 8 (citing, inter alia, Gideon v. Wainwright, 372 U. S. 335 (1963), and Turney v. Ohio, 273 U. S. 510 (1927)). The error at issue here is obviously quite different, however, as this Court’s post -Chapman decisions demonstrate. In Harrington v. California, for example, we expressly rejected the claim that the admission into evidence of a statement made by a nontestifying codefendant, in violation of Bruton v. United States, supra, can never be harmless.5 Harrington, which we have expressly reaffirmed on more than one occasion, see, e. g., Schneble v. Florida, supra; Brown v. United States, 411 U. S. 223 (1973), demonstrates that the denial of the opportunity to cross-examine an adverse witness does not fit within the limited category of constitutional errors that are deemed prejudicial in every case.

Respondent seeks to blunt the force of Harrington in essentially two ways. First, he suggests that this Court’s decision in Davis v. Alaska forecloses application of harmless-error analysis to the particular sort of Confrontation Clause violation involved here, citing the following language near the end of the Court’s opinion:

“[Davis] was thus denied 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. *6831, 3.’” 415 U. S., at 318 (quoting Smith v. Illinois, 390 U. S. 129, 131 (1968)).

Read properly, however, Davis does not support an automatic reversal rule, and the above-quoted language merely reflects the view that on the facts of that case the trial court’s error had done “serious damage” to the petitioner’s defense.

Davis was charged with stealing a safe from a bar. The police found the stolen safe abandoned near the home of Richard Green, who testified at trial that he had seen Davis engaged in suspicious activity near this site on the day of the crime. Defense counsel was barred from eliciting on cross-examination that Green was on juvenile probation for burglary both at the time of his pretrial identification of Davis and at the time of trial. The defense sought to suggest that Green may have slanted his account in the State’s favor either to shift suspicion away from himself or to avoid revocation of probation for failing to “cooperate.” 415 U. S., at 310-311. This Court reversed Davis’ conviction, emphasizing that Green’s tesimony was “crucial” and that there was a “real possibility” that pursuit of the excluded line of impeachment evidence would have done “[sjerious damage to the strength of the State’s case.” Id., at 319. So despite the absence of a reference to Chapman, Davis plainly rests on the conclusion that on the facts of that case, the error might well have contributed to the guilty verdict. Davis should not be read as establishing, without analysis, a categorical exception to the harmless-error rule.

Respondent’s second argument in support of a per se reversal rule is that the Confrontation Clause error in this case, which like Davis involved the exclusion of evidence, is analytically distinct from that in Harrington v. California, which involved the erroneous admission of harmless testimony. Because it is impossible to know how wrongfully excluded evidence would have affected the jury, the argument runs, reversal is mandated. But Harrington cannot be so easily dispatched. Respondent, like Harrington, was denied *684an opportunity to cast doubt on the testimony of an adverse witness.6 In both cases the prosecution was thus able to introduce evidence that was not subject to constitutionally adequate cross-examination. And in both cases the reviewing court should be able to decide whether the not-fully-impeached evidence might have affected the reliability of the factfinding process at trial.

Accordingly, we hold that the constitutionally improper denial of a defendant’s opportunity to impeach a witness for bias, like other Confrontation Clause errors, is subject to Chapman harmless-error analysis. The correct inquiry is whether, assuming that the damaging potential of the cross-examination were fully realized, a reviewing court might nonetheless say that the error was harmless beyond a reasonable doubt. Whether such an error is harmless in a particular case depends upon a host of factors, all readily accessible to reviewing courts. These factors include the importance of the witness’ testimony in the prosecution’s case, whether the testimony was cumulative, the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points, the extent of cross-examination otherwise permitted, and, of course, the overall strength of the prosecution’s case. Cf. Harrington, 395 U. S., at 254; Schneble v. Florida, 405 U. S., at 432.

We believe that the determination whether the Confrontation Clause error in this case was harmless beyond a reasonable doubt is best left to the Delaware Supreme Court in the first instance. Accordingly, that court’s judgment is vacated, and the case is remanded for further proceedings not inconsistent with this opinion.

It is so ordered.

concurring in the judgment.

The Sixth Amendment confers on defendants in criminal cases the right “to be confronted with the witnesses against” *685them. The Court has interpreted these words as meaning more than being allowed to confront the witnesses physically, more than the right to be tried by live testimony rather than affidavits. It includes the opportunity for effective cross-examination of the State’s witnesses. I do not here dispute these interpretations of the constitutional language; but they neither require nor advise the Court to hold, as it does today, that the Amendment is violated whenever a trial judge limits cross-examination of a particular witness and the jury might have received a significantly different impression of the witness’ credibility had cross-examination not been curtailed, even if the limitation and its consequences could not possibly have had any effect on the outcome of the trial.

It makes much more sense to hold that no violation of the Confrontation Clause has occurred unless there is some likelihood that the outcome of the trial was affected. I agree that the Delaware Court erred and that we should remand for consideration of prejudice, but I would not now hold that a constitutional violation occurred. If it is ultimately held that the outcome would have been the same whether or not cross-examination had been limited, no Sixth Amendment violation occurred in this case.

I would thus treat this claim of a Sixth Amendment violation just as the majority would treat limitations on cross-examination that would fall within the trial judge’s “wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on . . . cross-examination based on concerns about, among other things, harassment, prejudice, confusion of the issues, the witness’ safety, or interrogation that is repetitive or only marginally relevant.” Ante, at 679. These “reasonable” limitations are not violations at all, obviously because they can have no impact on the fairness of the trial. Yet the curtailment of cross-examination imposed in this case is said to be unreasonable and an infraction of the Amendment even though it may be held beyond reasonable doubt that it had no impact whatsoever on the result the jury reached.

*686No judge welcomes or can ignore being told that he committed a constitutional violation, even if the conviction is saved by a harmless-error finding. Being advised by the Court that there is an area of cross-examination curtailment that is not only harmless but not a constitutional violation but at the same time an area of curtailment that even though harmless is an infraction of our fundamental charter, the judge will surely tend to permit the examination rather than risk being guilty of misunderstanding the constitutional requirements of a fair trial. I would not so undermine the authority of the judge to restrict cross-examination in a manner having no appreciable impact on the reliability of the outcome, particularly since the language and purpose of the specific provision at issue do not otherwise dictate.

Even if it is ultimately held in this case that the error was harmless, as the Court is quite willing to assume will be the case, the judge has been declared derelict and commanded not again to restrict cross-examination in this manner even though he is convinced, and rightly so, that it has no significance whatsoever in terms of the outcome of the trial. With all due respect, I cannot join the Court’s opinion.

dissenting.

The Court today properly holds that a complete denial of cross-examination designed to explore the bias of a prosecution witness violates the Confrontation Clause, whether or not the denial influenced the outcome of the trial and whether or not the witness was important to the prosecution’s case. Nevertheless, the Court remands in order to permit the state court to apply harmless-error analysis to that violation. I must respectfully dissent from the latter part of the Court’s holding. I believe that the importance of cross-examination to a criminal trial is so great that a complete denial of otherwise proper cross-examination concerning the potential bias of a prosecution witness should lead to no less than a reversal of the conviction.

*687In holding the Confrontation Clause applicable to the States, this Court referred to the right of cross-examination as “an essential and fundamental requirement for the kind of fair trial which is this country's constitutional goal.” Pointer v. Texas, 380 U. S. 400, 405 (1965). If indeed the harmless-error doctrine focuses on the fairness and accuracy of a criminal trial, see ante, at 681, it is odd that the majority so easily applies it to a type of error that calls both fairness and accuracy into question to an almost unique degree.

The centrality of cross-examination to the factfinding process makes it particularly unlikely that an appellate court can determine that a denial of cross-examination had no effect on the outcome of a trial.

“[T]he court ordinarily cannot measure whether harm has ensued to an appellant when he has been denied the opportunity to cross-examine witnesses against him, given all the risks. Had cross-examination been allowed, for example, it might have served to impeach a witness and thus to cast doubt on corroborating testimony, or it might have elicited exculpatory evidence. Only on rare occasions will an appellate court be able to find that the testimony of the witness was so tangential, or so well corroborated, or so clearly invulnerable to attack that the denial of the right to cross-examination was harmless.” R. Traynor, The Riddle of Harmless Error 68-69 (1970).

The fact that a particular witness’ testimony was corroborated cannot render harmless a denial of the right to expose his bias. Defense counsel may have valid strategic reasons for challenging one witness’ testimony aggressively while treating a corroborating witness more gently. Jurors evaluating the witnesses’ demeanor may choose to give great weight to the testimony of one witness while ignoring the similar testimony of another. In either event, denial of cross-examination concerning a witness’ bias may deprive the defense of its best opportunity to expose genuine flaws in the *688prosecution’s case — flaws that the cold record will not reveal to an appellate court.

Indeed, an appellate court attempting to apply harmless-error analysis is faced with a formidable burden. The court cannot merely satisfy itself that the jury would have reached the same result had the witness in question not appeared at all; it must be convinced beyond a reasonable doubt that the jury would have reached the same result even if cross-examination had led the jury affirmatively to. believe that the witness was lying. Moreover, the court must conclude, beyond a reasonable doubt, that no evidence exculpatory to the defendant could have emerged from a genuinely adversarial testing of the witness. I think that a court can make such a determination only in the rarest of circumstances, and a rule of per se reversal is therefore justified.

The Confrontation Clause violation in this case is especially pernicious. The jury was essentially misled, by the empty gesture of cross-examination, to believe that the defense attorney had been permitted to use all the tools at his disposal to expose weaknesses in Fleetwood’s testimony. Having survived what appeared to be counsel’s best efforts to undermine the witness’ credibility, Fleetwood’s testimony necessarily carried more weight with the jury than would the same testimony given without an apparent opportunity to cross-examine.

This analysis makes it unnecessary to strain, as does the majority, to reconcile the apparent per se rule of Davis v. Alaska, 415 U. S. 308 (1974), with the harmless-error analysis employed in Harrington v. California, 395 U. S. 250 (1969), and Schneble v. Florida, 405 U. S. 427 (1972). I would simply hold that Davis mandates reversal whenever the prosecution puts a witness on the stand but the court does not- permit the defense to cross-examine concerning relevant potential bias. I therefore dissent from the Court’s decision to permit the Delaware Supreme Court to apply *689harmless-error analysis to the Confrontation Clause violation in this case.

I also write to emphasize that this Court cannot require state courts to apply harmless-error analysis to violations of the Federal Constitution. See Connecticut v. Johnson, 460 U. S. 73, 88 (1983) (Stevens, J., concurring in judgment). While this Court has stated that federal law governs the application of harmless error to violations of the Federal Constitution, see Chapman v. California, 386 U. S. 18, 21 (1967), that cannot mean more than that state courts must reverse convictions when the Constitution so mandates. When the Constitution does not mandate a particular remedy, this Court may not “declare which of many admittedly constitutional [remedial] alternatives a State may choose.” Id., at 48 (Harlan, J., dissenting) (footnote omitted). We have never held that the Federal Constitution forbids state courts to reverse certain convictions pursuant to state substantive or procedural law, nor can I imagine what provision of the Constitution could grant us such a power. Thus the Delaware Supreme Court remains free on remand to decide that even though it applied the substantive standards of the Sixth Amendment to determine whether error occurred, its harmless-error analysis was the product of state rather than federal law.

dissenting.

The Court finds the way open to reverse the judgment in this case because “[t]he opinion of the Delaware Supreme Court, which makes use of both federal and state cases in its analysis, lacks the requisite ‘plain statement’ that it rests on state grounds.” Ante, at 678, n. 3.1 In so holding, the Court *690continues down the path it marked in Michigan v. Long, 463 U. S. 1032, 1037-1044 (1983), when it announced that it would henceforth presume jurisdiction to review state-court judgments absent a “plain statement” that such judgments rest on state grounds.2

Despite the directness of the route chosen, today’s destination was not foreordained. Unlike Michigan v. Long, this case concerns whether the Court should presume jurisdiction to review a state supreme court’s remedy for a federal constitutional violation. Since courts have traditionally enjoyed broad discretion to fashion remedies — even remedies forbidding otherwise lawful acts — once a constitutional violation *691has been proved,3 the more logical direction would have been to presume that a state court is merely exercising its normal supervisory power over state officials unless it clearly states that federal law requires a particular procedure to be followed. The Court’s contrary presumption works a further advancement of its own power, but it flouts this Court’s best traditions: it deviates from our normal approach to questions of subject-matter jurisdiction, and it departs from our longstanding practice of reserving decision on federal constitutional law. Even considered purely from the standpoint of managing our own discretionary docket, the Court’s presumption includes a selection bias inconsistent with the lessons of history as revealed in this Court’s statutory jurisdiction over the judgments of state courts. Finally, the Court’s willingness to presume jurisdiction to review state remedies evidences a lack of respect for state courts and will, I fear, be a recurrent source of friction between the federal and state judiciaries.

I

The rules that govern this Court’s jurisdiction to review state-court judgments should, of course, be consistent with the jurisdictional principles that govern the entire federal ju*692dicial system. Indeed, because the example that this Court sets for the entire system inevitably affects the way in which all federal judges tend to evaluate their own powers, we have a special obligation to make sure that our conclusions concerning our own jurisdiction rest on a firm and legitimate foundation.

In origin and design, federal courts are courts of limited jurisdiction; they exercise only the authority conferred on them by Art. Ill and by congressional enactments pursuant thereto. See Bender v. Williamsport Area School Dist., ante, at 541, and cases cited therein. Like all other federal courts, this Court has only the power expressly given it. Because it is our inescapable duty — in contrast to that of the political branches — to construe authoritatively the very instruments which define and limit that power, the Court early in its history wisely adopted a presumption that every federal court is “without jurisdiction” unless “the contrary appears affirmatively from the record.” King Bridge Co. v. Otoe County, 120 U. S. 225, 226 (1887). Accord, Thomas v. Board of Trustees, 195 U. S. 207, 210 (1904); Minnesota v. Northern Securities Co., 194 U. S. 48, 62-63 (1904). That presumption is just as “inflexible” in this Court as in any other federal court.4

Even for cases unquestionably within this Court’s subject-matter jurisdiction, we have disclaimed any pretension to *693reach questions arising under the Federal Constitution when an alternative basis of decision fairly presented itself. Thus, in one of the most respected opinions ever written by a Member of this Court, Justice Brandéis wrote:

“The Court [has] developed, for its own governance in the cases confessedly within its jurisdiction, a series of rules under which it has avoided passing upon a large part of all the constitutional questions pressed upon it for decision. They are:

[[Image here]]

“. . . The Court will not pass upon a [federal] constitutional question although properly presented by the record, if there is also present some other ground upon which the case may be disposed of.” Ashwander v. TVA, 297 U. S. 288, 346-347 (1936) (concurring opinion).6

The Court has remained faithful to these basic tenets when it is reviewing cases that arise in the federal system. See Bender v. Williamsport Area School Dist., ante, at 545-549; Regents of University of Michigan v. Ewing, 474 U. S. 214, 222-223 (1985). Ironically, however — and contrary to tradition6 — the Court has taken a different stance when it is *694asked to review cases coming to us from state courts. Although “[w]e cannot perform our duty to refrain from interfering in state law questions and also to review federal ones without making a determination whether the one or the other controls the judgment,” Herb v. Pitcairn, 324 U. S. 117, 125-126 (1945), the jurisdictional precepts that serve us so well in reviewing judgments rendered in federal court merit observance in review of state-court judgments too. Abjuring the federal analogy, the Court unwisely marks for special scrutiny the decisions of courts to which I believe it owes special respect.

II

The jurisdictional presumption that the Court applies — and extends — today harbors a hidden selection bias that in turn reveals a disturbing conception of this Court’s role. Because a state ground can only support a judgment consistent with a *695federal claim, the Court’s jurisdictional presumption operates to expand this Court’s review of state remedies that overcompensate for violations of federal constitutional rights. Historically, however, such cases have been outside the province of this Court. For well over a century the Judiciary Act of 1789 denied this Court authority to review state-court judgments upholding federal claims.7 By conferring no power to review these judgments, “the first Congress assembled under the Constitution” — whose Members had “taken part in framing that instrument,” Wisconsin v. Pelican Ins. Co., 127 U. S. 265, 297 (1888), in addition to having enacted the First Judiciary Act — codified their conviction that this Court’s overriding concern was to ensure that state courts respect federal rights. Only in 1914 did Congress authorize this Court to take jurisdiction over state-court judgments upholding claims of federal constitutional right, Act of Dec. 23, 1914, 38 Stat. 790, and even that legislation reflected an un*696derstanding that the Court’s role is primarily to vindicate such rights.8 Most of the bills on this subject gave “the litigant an absolute right to appeal or take a writ of error to *697the Supreme Court. . . even though the decision is in favor of a claim of right under the Federal Constitution.” S. Rep. No. 161, 63d Cong., 2d Sess., 2 (1914). Rather than adopt these bills, which would have placed uniformity of federal law on a par with vindication of federal rights by making review of such judgments at least nominally mandatory, Congress “substitute^] a grant of jurisdiction to the Supreme Court of the United States to issue a writ of certiorari or otherwise to review the decision of the State court.” Ibid. Compare Act of Sept. 24,1789, § 25,1 Stat. 85-86, with Act of Dec. 23, 1914, 38 Stat. 790. Thus, although this Court now has the power to review decisions defending federal constitutional rights, the claim of these cases on our docket is secondary to the need to scrutinize judgments disparaging those rights.9 *698When the state-court decision to be reviewed is ambiguous, and it is not even clear that the judgment rests on a federal ground, the basis for exercising jurisdiction is even less tenable.

Ill

The Court’s decision to monitor state-court decisions that may or may not rest on nonfederal grounds is not only historically disfavored but risks the very confrontations and tensions a more humble jurisdictional stance would avoid. The presumption applied today allocates the risk of error in favor of the Court’s power of review; as a result, over the long run *699the Court will inevitably review judgments that in fact rest on adequate and independent state grounds. Even if the Court is unconcerned by the waste inherent in review of such cases, even if it is unmoved by the incongruity between the wholly precatory nature of our pronouncements on such occasions and Art. Ill’s prohibition of advisory opinions, it should be concerned by the inevitable intrusion upon the prerogatives of state courts that can only provide a potential source of friction and thereby threaten to undermine the respect on which we must depend for the faithful and conscientious application of this Court’s expositions of federal law.

Less obvious is the impact on mutual trust when the state court on remand — perhaps out of misplaced sense of duty— confines its state constitution to the boundaries marked by this Court for the Federal Constitution. In Montana v. Jackson, 460 U. S. 1030 (1983), for example, this Court vacated and remanded “for further consideration in light of South Dakota v. Neville, 459 U. S. 553 (1983).” In so doing, this Court presumed that the judgment of the Montana Supreme Court did not rest on Montana’s Constitution. Justice Sheehy, joined by the author of the state court’s original opinion, rather bitterly disagreed:

“In our original opinion in this case, we had examined the rights guaranteed our citizens under state constitutional principles, in the light of federal constitutional decisions. Now the United States Supreme Court has interjected itself, commanding us in effect to withdraw the constitutional rights which we felt we should extend to our state citizens back to the limits prescribed by the federal decisions. Effectively, the United States Supreme Court has intruded upon the rights of the judiciary of this sovereign state.

“Instead of knuckling under to this unjustified expansion of federal judicial power into the perimeters of our state power, we should show our judicial displeasure by insisting that in Montana, this sovereign state can inter*700pret its constitution to guarantee rights to its citizens greater than those guaranteed by the federal constitution.

[[Image here]]

“If a majority of this Court had the will to press the issue, we could put the question to the United States Supreme Court four-square, that this State judiciary has the right to interpret its constitution in the light of federal decisions, and to go beyond the federal decisions in granting and preserving rights to its citizens under its state constitution.” State v. Jackson, 206 Mont. 338, 349-351, 672 P. 2d 255, 260-261 (1983) (Sheehy, J., dissenting).

See id., at 357-358, 672 P. 2d, at 264-265 (Shea, J., dissenting).

The Court’s two-sentence analysis notwithstanding, one cannot be confident that we have not trenched on state prerogatives in this very case. Here, the Delaware Supreme Court applied a rule reversing convictions when the defendant had been totally denied the right to cross-examine a witness for bias. The rule was expressly found to be “consistent with Davis v. Alaska, 415 U. S. 308 (1974) and with our ruling in Weber [v. State, 457 A. 2d 674 (1983),] for determining whether a violation of the confrontation clause is harmless.” 486 A. 2d 1, 7 (1984) (emphasis added and citations omitted). Weber itself emphasized that “[b]oth the United States and Delaware Constitutions guarantee the right of a defendant to confront the witnesses against him. U. S. Const, amend. VI; Del. Const, art. I, § 7.” Weber v. State, 457 A. 2d, at 682 (footnote omitted). At no point did the Delaware Supreme Court imply that it reversed the defendant’s conviction only because that result was compelled by its understanding of federal constitutional law; rather, the conclusion that its rule was “consistent with” a case of this Court construing the federal Confrontation Clause suggests that it was interested merely in respecting the bounds of federal law as opposed to carrying out its command. The Court rewards *701the Delaware Supreme Court’s circumspection by unceremoniously reversing its judgment.

h-i <!

I agree with Justice Marshall that “the Delaware Supreme Court remains free on remand to decide that... its harmless-error analysis was the product of state rather than federal law.” Ante, at 689. Because the Court’s approach does nothing to minimize, and indeed multiplies, future occasions on which state courts may be called upon to clarify whether their judgments were in fact based on state law, it is appropriate to amplify the opinion I expressed in Massachusetts v. Upton, 466 U. S. 727, 736 (1984) (concurring in judgment), that the proper “sequence of analysis when an arguable violation of the State Constitution is disclosed by the record” is for the state court to consider the state constitutional claim in advance of any federal constitutional claim. In that case, I described the Oregon Supreme Court’s practice of considering state constitutional claims before reaching issues of federal constitutional law: *702Since that time, at least four other state courts have expressly endorsed the practice of considering state constitutional claims first.11 In response to Michigan v. Long, 463 *703U. S. 1032 (1983), for example, the New Hampshire Supreme Court concluded:

*701‘“The proper sequence is to analyze the state’s law, including its constitutional law, before reaching a federal constitutional claim. This is required, not for the sake either of parochialism or of style, but because the state does not deny any right claimed under the federal Constitution when the claim before the court in fact is fully met by state law.’ Sterling v. Cupp, 290 Ore. 611, 614, 625 P. 2d 123, 126 (1981).” Massachusetts v. Upton, 466 U. S., at 736.10

*703“When a defendant, as in this case, has invoked the protections of the New Hampshire Constitution, we will first address these claims.

*704“. . . We live under a unique concept of federalism and divided sovereignty between the nation and fifty States. The New Hampshire Constitution is the fundamental charter of our State. The sovereign people gave limited powers to the State government, and the Bill of Rights in part I of the New Hampshire Constitution protects the people from governmental excesses and potential abuses. When State constitutional issues have been raised, this court has a responsibility to make an independent determination of the protections afforded in the New Hampshire Constitution. If we ignore this duty, we fail to live up to our oath to defend our constitution and we help to destroy the federalism that must be so carefully safeguarded by our people. The Supreme Court of the State of Oregon recently recognized this responsibility and stated:

“ ‘The point is . .. that a state’s constitutional guarantees . . . were meant to be and remain genuine guarantees against misuse of the state’s governmental powers, truly independent of the rising and falling tides of federal case law both in method and specifics. State courts cannot abdicate their responsibility for these independent guarantees, at least not unless the people of the State themselves choose to abandon them and entrust their rights entirely to federal law.’ State v. Kennedy, 295 Or. 260, 271, 666 P. 2d 1316, 1323 (1983).” State v. Ball, 124 N. H. 226, 231, 471 A. 2d 347, 350 (1983).

Since 1983, in over a dozen cases,12 the New Hampshire Supreme Court has thereby averted unnecessary disquisitions on the meaning of the Federal Constitution.

*705The emerging preference for state constitutional bases of decision in lieu of federal ones is, in my view, the analytical approach best suited to facilitating the independent role of state constitutions and state courts in our federal system. There is much wisdom in The Chief Justice’s admonition that.“State courts . . . are responsible first for resolving issues arising under their constitutions and statutes and then for passing on matters concerning federal law.” Year-End Report on the Judiciary 18 (1981).

It must be remembered that every State but Rhode Island had a written constitution by the close of the Revolutionary War in 1783. “[F]or the first century of this Nation’s history, the Bill of Rights of the Constitution of the United States was solely a protection for the individual in relation to federal authorities. State Constitutions protected the liberties of the people of the several States from abuse by state authorities.” Massachusetts v. Upton, 466 U. S., at 738-739 (Stevens, J., concurring in judgment). The independent significance of state constitutions clearly informed this Court’s conclusion, in Barron v. The Mayor and City Council of Baltimore, 7 Pet. 243, 247-248 (1833), that the Bill of Rights applied only to the Federal Government:

“The question thus presented is, we think, of great importance, but not of much difficulty.

“The constitution was ordained and established by the people of the United States for themselves, for their own *706government, and not for the government of the individual states. Each state established a constitution for itself, and, in that constitution, provided such limitations and restrictions on the powers of its particular government as its judgment dictated.

“. . . In their several constitutions they have imposed such restrictions on their respective governments as their own wisdom suggested; such as they deemed most proper for themselves. It is a subject on which they judge exclusively, and with which others interfere no farther than they are supposed to have a common interest.”

While the holding of the Barron case has since been superseded by ratification of the Fourteenth Amendment and selective incorporation of the Bill of Rights, the concomitant atrophy of state constitutional theory was both unnecessary and unfortunate.13 State constitutions preceded the Federal Constitution and were obviously intended to have independent significance.14 The frequent amendments to state *707constitutions likewise presuppose their continued importance. Thus, whether the national minimum set by the Federal Constitution is high or low, state constitutions have their own unique origins, history, language, and structure — all of which warrant independent attention and elucidation. State courts remain primarily responsible for reviewing the conduct of their own executive branches, for safeguarding the rights of their citizenry, and for nurturing the jurisprudence of state constitutional rights which it is their exclusive province to expound.15

*708Because I would not presume that the Delaware Supreme Court failed to discharge this responsibility, I would dismiss the writ.