dissenting.
In this case, the District Court and the Court of Appeals for the Sixth Circuit ordered that respondent either be released or given a new hearing on the degree of his crimes and a resentencing. This was ordered despite the passage of 25 years since respondent’s convictions for two murders committed while he was a fugitive on escape from prison. Both courts held that laches did not bar respondent’s claim that he did not knowingly and intelligently waive his right to counsel at the hearing and sentencing in 1956. Understandably troubled by the possible ramifications of such a drastic holding and concerned that even in this particular case it would be prejudiced in its defense to the allegations, given the loss of records, faded memories, and intervening deaths, the State of Michigan sought certiorari to review the judgment of the Court of Appeals. We granted the State’s petition and set the case for argument. 469 U. S. 1033 (1984).
*155Now, having been informed that the respondent has died, the Court simply directs that our order granting certiorari be vacated and the petition for certiorari dismissed, thereby leaving the Court of Appeals’ opinion standing. In reaching this surprising result, the Court relies upon a single authority which, it is clear upon analysis, does not support, let alone require, such a disposition, see Warden v. Palermo, 431 U. S. 911 (1977). And it ignores the one precedent which does directly control on this question. See McMann v. Ross, 396 U. S. 118 (1969) (per curiam); McMann v. Richardson, 397 U. S. 759, 760, n. 1 (1970).
In McMann v. Ross, the only case to have presented the precise issue we have here, the Court vacated the judgment of the Court of Appeals and remanded to the District Court with instructions to dismiss the respondent’s petition for writ of habeas corpus as moot. I would, as petitioner urges, dispose of this case in the same way. This is the course we have chosen to pursue in every civil case that becomes moot either pending a decision on certiorari or after we have granted a writ of certiorari, except Warden v. Palermo, which even if it were correct is plainly distinguishable. Thirty-five years ago, the Court noted that
“[t]he established practice of the Court in dealing with a civil case from a court in the federal system which has become moot while on its way here or pending our decision on the merits is to reverse or vacate the judgment below and remand with a direction to dismiss.” United States v. Munsingwear, Inc., 340 U. S. 36, 39 (1950) (footnote omitted; emphasis added).
Indeed, as the Court said in Duke Power Co. v. Greenwood County, 299 U. S. 259, 267 (1936), this is the “duty” of this Court. Such a disposition “clears the path for future reliti-gation of the issues between the parties.” Munsingwear, supra, at 40. In this case it is possible, for example, that by applying collateral estoppel offensively, relatives of the defendant might well invoke the Sixth Circuit’s decision in a *156subsequent civil suit for damages. Moreover, by vacating the Court of Appeals’ judgment and remanding with instructions to dismiss as moot, we “eliminat[e] a judgment, review of which was prevented through happenstance.” 340 U. S., at 40. And it is only through this procedure that “the rights of all parties are preserved.” Ibid.
The Court mistakenly believes that our four-line order in Palermo, supra, requires that we only vacate the order granting certiorari and dismiss the petition. I do not understand how Palermo can possibly be regarded as controlling the disposition here. Palermo is significantly different from this case in at least one obvious respect — there, we had not granted the petition for certiorari. So also was the case in Dove v. United States, 423 U. S. 325 (1976), the authority we relied upon for our dismissal of the petition in Palermo. While in some circumstances we may wish to treat cases in which we have granted certiorari similarly to those in which a petition for certiorari has merely been filed here, it is inconceivable to me that we would ever consider ourselves bound to treat these two patently different categories of cases identically.
In my view, the Court has a higher duty when it learns that a case has become moot after it has granted review than when it discovers that a case in which review is being sought has become moot. In the former instance, the Court has, by its grant of the writ, asserted jurisdiction over the case, and brought the judgment of the lower court before it. If the Court simply vacates its order granting certiorari, it forces upon the parties and the courts below the problems it avoids for the sake of its own convenience. The Court’s action today leaves unclear, for instance, whether the Court of Appeals’ opinion remains as precedent as between the parties or their successors in any future proceeding. It is likewise unclear whether the opinion is generally to have precedential value in the Sixth Circuit. Finally, there remains considerable uncertainty over whether the Court of Appeals is *157required or permitted to vacate its opinion. Indeed, I suppose there will be a question whether the Court of Appeals even has jurisdiction to vacate or otherwise modify its opinion, given that our writ of certiorari is still lodged in that court; at the very least, the Court should vacate its writ of certiorari. Cf. Westinghouse Electric Corp. v. Vaughn, 466 U. S. 521 (1984); Colorado v. Nunez, 465 U. S. 324 (1984); Gillette Co. v. Miner, 459 U. S. 86 (1982). The Court’s disposition leaves the status of the Court of Appeals’ judgment and opinion in limbo. I believe we have an institutional obligation to avoid such confusion. This is easily achieved by following what heretofore, with the exception of Warden v. Palermo, has been our “established practice.”
Even if one accepts the dubious proposition that the Court is duty-bound to treat granted cases identical to cases where the petition for a writ of certiorari is pending, I still believe the Court’s disposition is incorrect because I am convinced the Court disposed of Palermo improperly. Palermo, it is true, was a case before us on habeas and we did dismiss the certiorari petition. The Court relied entirely, however, on Dove, a case which was before us on direct review, not ha-beas. In a case on direct review, it may be necessary simply to dismiss the petition to avoid the result in Durham v. United States, 401 U. S. 481 (1971), of having a defendant’s indictment dismissed, which in turn has the effect of “wip[ing] the slate entirely clean of a federal conviction which was unsuccessfully appealed throughout the entire appeal process to which the petitioner was entitled as of right,” id., at 484 (Blackmun, J., dissenting). The only alternative — and surely an unsatisfying one — would be for the Court to decide case-by-case whether it believes the decision in question to be correct or incorrect, and dismiss the petition for certiorari or vacate the judgment accordingly.
But plainly there is no such dilemma presented when a case is before us on writ of certiorari to review a judgment obtained on habeas. In such a case, our objectives of elimi*158nating as precedent an opinion and judgment of which no final review is possible and clearing the path for any future litigation are achievable — and incidentally, without at the same time embracing a principle that would require dismissal of indictments — by vacating the judgment below and remanding with instructions to dismiss the habeas petition.
Even the Court in Durham, recognized the validity of distinguishing in this context between cases on direct and habeas review; the Court very carefully limited its holding to cases on direct review, see id., at 482-483. Our order in Dove also contemplated this distinction. In Dove, we overruled Durham only “[t]o the extent that [Durham] may be inconsistent with” our disposition in Dove, 423 U. S., at 325. We thereby removed any doubt that Ross — which otherwise one might have thought the Court also intended to overrule — was still valid precedent. Under the circumstances, especially since Palermo not only relied upon inap-posite authority but failed even to acknowledge Ross, I would not, as the Court does, read Palermo as limiting us to a dismissal.
If it were true, however, as the Court implicitly must believe, that we are required now to overrule either Ross or Palermo, I would “overrule” the latter. Palermo is the case inconsistent with our asserted “established practice.” Palermo, not Ross, is the disposition in search of a rationale.
Because I believe we should not compound the evils of a bad practice by repeating the error here, I dissent.