with whom Justice Thomas joins, concurring in part and dissenting in part.
The issue in this case — whether the extraordinary remedy of federal habeas corpus should routinely be available for claimed violations of Miranda rights — involves not jurisdiction to issue the writ, but the equity of doing so. In my view, both the Court and Justice O’Connor disregard the most powerful equitable consideration: that Williams has already had full and fair opportunity to litigate this claim. He had the opportunity to raise it in the Michigan trial court; he did so and lost. He had the opportunity to seek review of the trial court’s judgment in the Michigan Court of Appeals; he did so and lost. Finally, he had the opportunity to seek discretionary review of that Court of Appeals judgment in both the Michigan Supreme Court and this Court; he did so and review was denied. The question at this stage is whether, given all that, a federal habeas court should now reopen the issue and adjudicate the Miranda claim anew. The answer seems to me obvious: it should not. That would be the course followed by a federal habeas court reviewing a federal conviction; it mocks our federal system to accord state convictions less respect.
I
By statute, a federal habeas court has jurisdiction over any claim that a prisoner is “in custody in violation of the Constitution or laws” of the United States. See 28 U. S. C. §§ 2241(c)(3), 2254(a), 2255. While that jurisdiction does require a claim of legal error in the original proceedings, cf. Herrera v. Collins, 506 U. S. 390 (1993), it is otherwise sweeping in its breadth. As early as 1868, this Court described it in these terms:
“This legislation is of the most comprehensive character. It brings within the habeas corpus jurisdiction of every court and of every judge every possible case of privation of liberty contrary to the National Constitu*716tion, treaties, or laws. It is impossible to widen this jurisdiction.” Ex parte McCardle, 6 Wall. 318, 325-326 (1868).
Our later case law has confirmed that assessment. Habeas jurisdiction extends, we have held, to federal claims for which an opportunity for full and fair litigation has already been provided in state or federal court, see Brown v.. Allen, 344 U. S. 443, 458-459 (1953); Kaufman v. United States, 394 U. S. 217, 223-224 (1969), to procedurally defaulted federal claims, including those over which this Court would have no jurisdiction on direct review, see Fay v. Noia, 372 U. S. 391, 426, 428-429 (1963); Kaufman, supra, at 223; Wainwright v. Sykes, 433 U. S. 72, 90-91 (1977); Coleman v. Thompson, 501 U. S. 722, 750 (1991), and to federal claims of a state criminal defendant awaiting trial, see Ex parte Royall, 117 U. S. 241, 251 (1886).
But with great power comes great responsibility. Habeas jurisdiction is tempered by the restraints that accompany the exercise of equitable discretion. This is evident from the text of the federal habeas statute, which provides that writs of habeas corpus “may be granted” — not that they shall be granted — and enjoins the court to “dispose of the matter as law and justice require.” 28 U. S. C. §§ 2241(a), 2243 (emphases added). That acknowledgment of discretion is merely the continuation of a long historic tradition. In English law, habeas corpus was one of the so-called “prerogative” writs, which included the writs of mandamus, certio-rari, and prohibition. Duker, The English Origins of the Writ of Habeas Corpus: A Peculiar Path to Fame, 53 N. Y. U. L. Rev. 983, 984, n. 2 (1978); 3 W. Blackstone, Commentaries 132 (1768). “[A]s in the case of all other prerogative writs,” habeas would not issue “as of mere course,” but rather required a showing “why the extraordinary power of the crown is called in to the party’s assistance.” Ibid. And even where the writ was issued to compel production of the *717prisoner in court, the standard applied to determine whether relief would be accorded was equitable: The court was to “determine whether the case of [the prisoner’s] commitment be just, and thereupon do as to justice shall appertain.” 1 id., at 131.
This Court has frequently rested its habeas decisions on equitable principles. In one of the earliest federal habeas cases, Ex parte Watkins, 3 Pet. 193, 201 (1830), Chief Justice Marshall wrote: “No doubt exists respecting the power [of the Court to issue the writ]; the question is, whether this be a case in which it ought to be exercised.” And in Ex parte Royall, the Court, while affirming that a federal habeas court had “the power” to discharge a state prisoner awaiting trial, held that it was “not bound in every case to exercise such a power.” 117 U. S., at 251. The federal habeas statute did “not deprive the court of discretion,” which “should be exercised in the light of the relations existing, under our system of government, between the judicial tribunals of the Union and of the States.” Ibid.
This doctrine continues to be reflected in our modern cases. In declining to extend habeas relief to all cases of state procedural default, the Court in Fay v. Noia said: “Discretion is implicit in the statutory command that the judge . . . ‘dispose of the matter as law and justice require,’ 28 U. S. C. § 2243; and discretion was the flexible concept employed by the federal courts in developing the exhaustion rule.” 372 U. S., at 438. See also Wainwright v. Sykes, supra, at 88. In fashioning this Court’s retroactivity doctrine, the plurality in Teague v. Lane, 489 U. S. 288, 308-310 (1989), also relied on equitable considerations. And in a case announced today, holding that the harmless-error standard for habeas corpus is less onerous than the one for direct ieview, the Court carries on this tradition by expressly considering equitable principles such as “finality,” “comity,” and “federalism.” Brecht v. Abrahamson, ante, at 635-636.
*718Indeed, as Justice O’Connor notes, this Court’s jurisprudence has defined the scope of habeas corpus largely by means of such equitable principles. See ante, at 698-700 (opinion concurring in part and dissenting in part). The use of these principles, which serve as “gateway[s]” through which a habeas petitioner must pass before proceeding to the merits of a constitutional claim, “is grounded in the 'equitable discretion’ of habeas courts.” Herrera v. Collins, supra, at 404.
II
As the Court today acknowledges, see ante, at 686-687, the rule of Stone v. Powell, 428 U. S. 466 (1976), is simply one application of equitable discretion. It does not deny a federal habeas court jurisdiction over Fourth Amendment claims, but merely holds that the court ought not to entertain them when the petitioner has already had an opportunity to litigate them fully and fairly. See id., at 495, n. 37. It is therefore not correct to say that applying Stone to the present case involves “eliminating review of Miranda claims” from federal habeas, ante, at 693, or that the Court is being “asked to exclude a substantive category of issues from relit-igation on habeas,” ante, at 700 (O’Connor, J., concurring in part and dissenting in part). And it is therefore unnecessary to discuss at length the value of Miranda rights, as though it has been proposed that since they are particularly worthless they deserve specially disfavored treatment. The proposed rule would treat Miranda claims no differently from all other claims, taking account of all equitable factors, including the opportunity for full and fair litigation, in determining whether to provide habeas review. Wherein Miranda and Fourth Amendment claims differ from some other claims, is that the most significant countervailing equitable factor (possibility that the assigned error produced the conviction of an innocent person) will ordinarily not exist.
At common law, the opportunity for full and fair litigation of an issue at trial and (if available) direct appeal was not *719only a factor weighing against reaching the merits of an issue on habeas; it was a conclusive factor, unless the issue was a legal issue going to the jurisdiction of the trial court. See Ex parte Watkins, supra, at 202-203; W. Church, Habeas Corpus §363 (1884). Beginning in the late 19th century, however, that rule was gradually relaxed, by the device of holding that various illegalities deprived the trial court of jurisdiction. See, e. g., Ex parte Lange, 18 Wall. 163, 176 (1874) (no jurisdiction to impose second sentence in violation of Double Jeopardy Clause); Ex parte Siebold, 100 U. S. 371, 376-377 (1880) (no jurisdiction to try defendant for violation of unconstitutional statute); Frank v. Mangum, 237 U. S. 309 (1915) (no jurisdiction to conduct trial in atmosphere of mob domination); Moore v. Dempsey, 261 U. S. 86 (1923) (same); Johnson v. Zerbst, 304 U. S. 458, 468 (1938) (no jurisdiction to conduct trial that violated defendant’s Sixth Amendment right to counsel). See generally Wright v. West, 505 U. S. 277, 285-286 (1992) (opinion of Thomas, J.); Fay, supra, at 450-451 (Harlan, J., dissenting). Finally, the jurisdictional line was openly abandoned in Waley v. Johnston, 316 U. S. 101, 104-105 (1942). See P. Bator, D. Meltzer, P. Mishkin, & D. Shapiro, Hart and Wechsler’s The Federal Courts and the Federal System 1502 (3d ed. 1988) (hereinafter Hart and Wechsler).
But to say that prior opportunity for full and fair litigation no longer automatically precludes from consideration even nonjurisdictional issues is not to say that such prior opportunity is no longer a relevant equitable factor. Reason would suggest that it must be, and Stone v. Powell, supra, establishes that it is. Thus, the question before us is not whether a holding unique to Fourth Amendment claims (and resting upon nothing more principled than our estimation that Fourth Amendment exclusion claims are not very important) should be expanded to some other arbitrary category beyond that; but rather, whether the general principle that is the only valid justification for Stone v. Poioell should for some *720reason not be applied to Miranda claims. I think the answer to that question is clear: Prior opportunity to litigate an issue should be an important equitable consideration in any habeas case, and should ordinarily preclude the court from reaching the merits of a claim, unless it goes to the fairness of the trial process or to the accuracy of the ultimate result.
Our case law since Stone is entirely consistent with this view. As the Court notes, ante, at 687-688, we have held that the rule in Stone does not apply in three cases. Kimmel-man v. Morrison, 477 U. S. 365 (1986), involved alleged denial of the Sixth Amendment right to counsel, which unquestionably goes to the fairness of the trial process. Rose v. Mitchell, 443 U. S. 545 (1979), involved alleged discrimination by the trial court in violation of the Fourteenth Amendment. We concluded that since the “same trial court will be the court that initially must decide the merits of such a claim,” and since the claim involved an assertion that “the state judiciary itself has purposely violated the Equal Protection Clause,” no opportunity for a full and fair state hearing existed. Id., at 561; see also id., at 563. And Jackson v. Virginia, 443 U. S. 307 (1979), involved a claim that “no rational trier of fact could have found proof of guilt beyond a reasonable doubt,” id., at 324, which is obviously a direct challenge to the accuracy of the ultimate result.
III
The rule described above — or indeed a rule even somewhat more limiting of habeas review than that — is followed in federal postconviction review of federal convictions under 28 U. S. C. § 2255. In Kaufman v. United States, 394 U. S. 217 (1969), which held that res judicata does not bar § 2255 habeas review of constitutional issues, we stated that a district court had “discretion” to refuse to reach the merits of a constitutional claim that had already been raised and resolved against the prisoner at trial and on direct review. *721Id., at 227, n. 8. Since Kaufman, federal courts have uniformly held that, absent countervailing considerations, district courts may refuse to reach the merits of a constitutional claim previously raised and rejected on direct appeal. See, e. g., Giacalone v. United States, 739 F. 2d 40, 42-43 (CA2 1984); United States v. Orejuela, 639 F. 2d 1055, 1057 (CA3 1981); Stephan v. United States, 496 F. 2d 527, 528-529 (CA6 1974), cert. denied sub nom. Marchesani v. United States, 423 U. S. 861 (1975); see also 3 C. Wright, Federal Practice and Procedure § 593, p. 439, n. 26 (1982); Note, Developments in the Law — Federal Habeas Corpus, 83 Harv. L. Rev. 1038, 1064-1066 (1970). Thus, a prior opportunity for full and fair litigation is normally dispositive of a federal prisoner’s ha-beas claim. If the claim was raised and rejected on direct review, the habeas court will not readjudicate it absent countervailing equitable considerations; if the claim was not raised, it is procedurally defaulted and the habeas court will not adjudicate it absent countervailing equitable considerations (e. g., actual innocence or cause and prejudice, see United States v. Frady, 456 U. S. 152 (1982)).
Because lower federal courts have not generally recognized their discretion to deny habeas relief in state cases where opportunity for full and fair litigation was accorded, the peculiar state of current federal habeas practice is this: State courts routinely see their criminal convictions vacated by federal district judges, but federal courts see their criminal convictions afforded a substantial measure of finality and respect. See Hart and Wechsler 1585. Only one theory can possibly justify this disparity — the theory advanced in Fay v. Noia, that a federal forum must be afforded for every federal claim of a state criminal defendant* See 372 U. S., at *722418. In my view, that theory is profoundly wrong for several reasons.
First, it has its origin in a misreading of our early precedents. Fay interpreted the holding of Ex parte Royall— that federal courts had discretion not to entertain the habeas claims of state prisoners prior to the conclusion of state-court proceedings — as containing the implication that after conclusion of those proceedings there would be plenary federal review of all constitutional claims. 372 U. S., at 420. In fact, however, Royall had noted and affirmed the common-law rule that claims of error not going to the jurisdiction of the convicting court could ordinarily be entertained only on writ of error, not on habeas corpus. 117 U. S., at 253. See Fay, supra, at 453-454 (Harlan, J., dissenting). See also Sch-neckloth v. Bustamonte, 412 U. S. 218, 255 (1973) (Powell, J., concurring). Royall contained no hint of a suggestion that a federal habeas court should afford state-court judgments less respect than federal-court judgments. To the contrary, it maintained the traditional view that federal and state courts have equal responsibility for the protection of federal constitutional rights. The discretion of the federal habeas court “should be exercised,” it said, “in the light of the relations existing, under our system of government, between the judicial tribunals of the Union and of the States, . . . courts equally bound to guard and protect rights secured by the Constitution.” 117 U. S., at 251. And in describing the proper disposition of a federal habeas petition filed after state conviction, Royall cited Ex parte Lange, 18 Wall. 163 (1874), which involved a federal habeas attack on a federal conviction. See 117 U. S., at 253. Thus, Royall is properly understood as saying that the federal habeas statute guaranteed state prisoners, not a federal forum for all their federal claims, but rather the same rights to federal habeas relief that federal prisoners possessed.
Worse than misreading case precedent, however, the federal right/federal forum theory misperceives the basic struc*723ture of our national system. That structure establishes this Court as the supreme judicial interpreter of the Federal Constitution and laws, but gives other federal courts no higher or more respected a role than state courts in applying that “Law of the Land” — which it says all state courts are bound by, and all state judges must be sworn to uphold. U. S. Const., Art. VI. See Robb v. Connolly, 111 U. S. 624, 637 (1884); Ex parte Royall, supra, at 251; Brown, 344 U. S., at 499 (opinion of Frankfurter, J.). It would be a strange constitution that regards state courts as second-rate instruments for the vindication of federal rights and yet makes no mandatory provision for lower federal courts (as our Constitution does not). And it would be an unworkable constitution that requires redetermination in federal courts of all issues of pervasive federal constitutional law that arise in state-court litigation.
Absent indication to the contrary, state courts should be presumed to have applied federal law as faithfully as federal courts. See Ex parte Royall, supra, at 252; Brecht v. Abrahamson, ante, at 636. A federal court entertaining collateral attack against a state criminal conviction should accord the same measure of respect and finality as it would to a federal criminal conviction. As it exercises equitable discretion to determine whether the merits of constitutional claims will be reached in the one, it should exercise a similar discretion for the other. The distinction that has arisen in lower court practice is unsupported in law, utterly impractical and demeaning to the States in its consequences, and must be eliminated.
* * *
While I concur in Part III of the Court’s opinion, I cannot agree with the rest of its analysis. I would reverse the judgment of the Court of Appeals and remand the case for a determination whether, given that respondent has already been afforded an opportunity for full and fair litigation in the *724courts of Michigan, any unusual equitable factors counsel in favor of readjudicating the merits of his Miranda claim on habeas corpus.
Of course a federal forum is theoretically available in this Court, by writ of certiorari. Quite obviously, however, this mode of review cannot be generally applied due to practical limitations. See Stone v. Powell, 428 U. S. 465, 526 (1976) (Brennan, J., dissenting).