I part ways with the majority over what may appear to be a small point, but one *1166which I believe has large implications. In my view, there is no final order in this case, as required by 28 U.S.C. § 1291, and thus we lack appellate jurisdiction.
I.
The rule that appeals lie only from final judgments is not only “[t]he historic rule in the federal courts,” Advisory Committee Note of 1946 to amended Rule 54(b), Fed.R. Civ.P., but represents a considered policy judgment on the part of Congress that has been emphasized repeatedly by the federal judiciary. See, e.g., Catlin v. United States, 324 U.S. 229, 233, 65 S.Ct. 631, 634, 89 L.Ed. 911 (1945) (“The foundation of the policy is not in merely technical conceptions of ‘finality.’ It is one against piecemeal litigation.”). As Justice Frankfurter observed, the final judgment rule
was written into the first Judiciary Act and has been departed from only when observance of it would practically defeat the right to any review at all____ Congress from the very beginning has, by forbidding piecemeal disposition on appeal of what for practical purposes is a single controversy, set itself against enfeebling judicial administration. Thereby is avoided the obstruction to just claims that would come from permitting the harassment and cost of a succession of separate appeals from the various rulings to which a litigation may give rise, from its initiation to entry of judg-ment____
Cobbledick v. United States, 309 U.S. 323. 324-25, 60 S.Ct. 540, 541, 84 L.Ed. 783 (1940). Thus, this Court itself has stated:
The substance of the dispute here is highly significant, and immediate resolution would clarify an important aspect of ... law as well as possibly terminating a lengthy controversy. However, the wisdom of the final judgment rule lies in its insistence that we focus on systemic, as well as particularistic impacts. The appellate system has become increasingly overburdened and the future would appear to promise no relief from the continuous increase in case loads. Accordingly, it would seem to us to be a disservice to the Court, to litigants in general and to the idea of speedy justice if we were to succumb to enticing suggestions to abandon the deeply-held distaste for piecemeal litigation in every instance of temptation.
Bachowski v. Usery, 545 F.2d 363, 373-74 (3d Cir.1976). This very language and reasoning were adopted in Hoots v. Commonwealth of Pennsylvania, 587 F.2d 1340, 1347 n. 40 (3d Cir.1978).
II.
“In a multiparty lawsuit such as this one, an order is final only if it meets one or the other of two conditions: (1) it must adjudicate the claims or the rights and liabilities of all the parties, or (2) [the district court] must expressly determine that there is no just reason for delay and expressly direct the entry of judgment. Rule 54(b), F.R. Civ.P.” Jetco Electronic Industries, Inc. v. Gardiner, 473 F.2d 1228, 1231 (5th Cir. 1973). The only alternative route by which review may be sought of a non-final order is to obtain certification of the question to this Court by the district court. 28 U.S.C. § 1292(b).1 But such certification was expressly sought in this case, and was specifically denied by the district court.
I agree with the majority that “it would be anomalous to hold that a plaintiff had no right to appeal the dismissal of all but one of his claims after that one claim not initially dismissed, had thereafter been voluntarily and finally abandoned.” Ante, at 1155. *1167I also agree with the majority that “[ujnder Fed.R.App.P. 4(a)(2) and the Cape May Greene reasoning, the ‘premature’ notices of appeal filed on October 15, 1985 would have become effective upon the final abandonment of the claims against Troy, and appellate jurisdiction would then have vested.” The problem here, however, is that the remaining claim — that against Christopher Troy — has not been “finally abandoned.”
The majority maintains that the order of voluntary dismissal without prejudice represents a final order because it has been more than two years since the accident occurred, and because the plaintiffs themselves claimed at oral argument that they have abandoned their right to press further their claim against Troy. But these arguments are not sustainable on the facts before us.
Although the two years normally allotted for the bringing of actions under Pennsylvania’s statute of limitations had already passed as of the time of Troy’s dismissal without prejudice, the statute of limitations issue is neither properly before us, nor dispositive of this case. The statute of limitations was not pleaded, and was not raised in the district court, or in the briefs filed in this Court. A statute of limitations time bar is not jurisdictional; rather, it constitutes an affirmative defense that is waived if the defendant fails to raise it in the answer. Fed.R.Civ.P. 8(c), 12(c). This Court therefore may not address the issue sua sponte. The fact that the statute of limitations could not have been raised at any time in the history of this case to date, as the majority notes, is exactly the point. We cannot predict whether Troy will assert the time bar until he has an opportunity to do so. Until that occurs — when the plaintiffs attempt to reinstitute the action against Troy, as permitted under the district court’s order — it is merely speculative whether the plaintiffs will be unable to proceed. The majority’s approach is simply not an instance of an appellate court assuming its duty to decline jurisdiction where some fact exists barring such jurisdiction; rather, the majority here attempts to assert jurisdiction where there is some possibility that a future event might eliminate the jurisdictional bar.
Moreover, on the record before us, there is nothing to indicate that the action against Troy may not be reinstituted by Fassett following an unfavorable decision on this appeal. Carr v. Grace, 516 F.2d 502 (5th Cir.1975), which the majority cites for the proposition that “the running of the statute of limitations ma[kes] a ‘without prejudice’ dismissal final for purposes of 28 U.S.C. § 1291,” ante at 1155, confined this ruling to “the peculiar circumstances of that case,” and noted that “[t]he appealability of an order depends on its effect rather than its language.” Carr, 516 F.2d at 503 n. 1. Neither the effect nor the apparent intention of the parties here was that the dismissal without prejudice would serve to foreclose further proceedings against Troy. Indeed, the colloquy on this point at oral argument suggests that, in fact, the intention of the parties was to obtain a de facto interlocutory ruling at this time and to preserve the plaintiffs’ option to proceed against Troy thereafter (whether or not it would later appear to be remunerative to do so).
Except in the most unusual circumstances, factual representations advanced for the first time at oral argument, especially when unsupported by any affidavits, should not be addressed by this Court, as they do not constitute any part of the Record before us. As the majority does not refer in full to the oral argument, however, it is necessary to note in turn the remainder of counsels’ assertions at oral argument in order to present a complete picture. At oral argument counsel admitted that Troy, Fassett, and Buckley had all agreed that a de facto interlocutory appeal would be arranged, whereby the plaintiffs would voluntarily dismiss the claim against Troy, Troy would agree that the dismissal be “without prejudice,” and Troy would not assert the statute of limitations defense if Fassett and Buckley lost on appeal and then reinstituted the action against Troy. Given this admission by counsel, it *1168is somewhat disingenuous to “assume that Troy would fail to plead a statute of limitations defense only if he were unaware of its availability,” and to suggest that the possibility that Fassett and Buckley might reinstate their action and that Troy might not plead the statute of limitations is “mere[ly] theoretical” or “unrealistic.” Ante at 1156.
In short, the parties worked out the sort of “end-run” around the finality rule that the courts have always disallowed and that the majority’s approach today passively endorses. Allowing a party, by taking a voluntary dismissal without prejudice, to obtain a de facto interlocutory appeal — especially where certification under § 1292(b) has been denied — would undercut the statutes governing appealable orders and the underlying philosophy disfavoring piecemeal litigation. As the Ninth Circuit declared in Fletcher v. Gagosian, 604 F.2d 637, 639 (9th Cir.1979), “if we accept appellants’ rationale, then we also accept the notion that the policies against multiplicity of litigation and against piecemeal appeals may be avoided at the whim of a plaintiff. He need merely dismiss portions of his complaint without prejudice, appeal from what had been an interlocutory order, and refile the dismissed portion as a separate lawsuit.”
I believe, as does the majority, that we must “giv[e] a practical rather than a technical construction to the finality rule,” ante at 1156 (quoting Anderson v. Allstate Insurance Co., 630 F.2d 677, 681 (9th Cir. 1980). Unlike Anderson, however, in which finality was achieved after the filing of the appeal, we cannot say here that “nothing else remains in the federal courts.” Id. The claim against Troy may still be reinstated at any time, and the representations of counsel appear to make plain that the parties have reserved that possibility.2 The dismissal remains without prejudice, and until the defendant raises the limitations bar, the plaintiffs retain their cause of action under the dismissal order.
The parties themselves may pursue either of two solutions to the appealability problem. They may return to the district court and seek a dismissal with prejudice of the claim against Troy, ending the possibility and suspicion that they have entered into a mutually agreed upon subterfuge to obtain interlocutory review while keeping alive the possibility of an action against Troy. Or they may proceed against Troy — a trial that would likely be quite brief — and then, if appropriate, seek proper review of the final judgment. One or the other of these decisions might affect the possible state proceedings against Troy, but that is a result of the two-track strategy the plaintiffs have elected to pursue. In any case, if the majority believes that counsel meant what they said (at least as their assertions at oral argument have been reconstructed), then it is difficult to see how any of these possible outcomes prejudices the plaintiffs in the least, since “as a pragmatic matter” the plaintiffs would not proceed against Troy given that litigation costs would outweigh the limited insurance yield that Troy represents. See ante at 1156-1157.
Whether we look exclusively to the Record — from which we see only that the cause was dismissed without prejudice — or to the representations of counsel at oral argument — from which we learn that the dismissal without prejudice was part of an arrangement that included Troy’s promise not to raise the statute of limitations — it is clear that this is not a case of relying on a *1169narrow technicality versus upholding the spirit of the law. Rather this is a case of upholding the spirit of the law versus an attempt to undercut the rule of law by artifice. See, e.g., Sullivan v. Pacific Indemnity Co., 566 F.2d 444, 445 (3d Cir. 1977) (noting “our disapproval of indirect attempts to accomplish what cannot be done directly”). I sympathize with the concern that cases not be shuttled needlessly back and forth between the district courts and the courts of appeals, but the problem here has been created by the parties themselves.
The result reached by the majority today creates a loophole in our jurisdictional limitations, opening the door for future erosion of the principle behind the final judgment rule. Such erosion can only cause greater injustice, inefficiency, and delay in the future: “To review the district court’s [decision] under the facts of this case is ... to undermine the ability of trial judges to achieve the orderly and expeditious disposition of cases.” Marshall v. Sielaff, 492 F.2d 917, 919 (3d Cir.1974).3
III.
I would conclude that this Court lacks jurisdiction to consider the case at this time, and that the only appropriate course is to dismiss the appeal. Accordingly, I respectfully dissent.
. The doctrine of Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541, 69 S.Ct. 1221, 93 L.Ed. 1528 (1949), allows appellate review of collateral orders. For an otherwise unappealable order to be deemed "final" under the collateral order doctrine of Cohen, "the order must conclusively determine the disputed question, resolve an important issue completely separate from the merits of the action, and be effectively unreviewable on appeal from a final judgment." Yakowicz v. Commonwealth of Pennsylvania, 683 F.2d 778, 783 (3d Cir.1982). The Cohen doctrine thus clearly does not apply to this case. The issue on appeal is not collateral, nor is it one that would evade subsequent review.
. Even if it were true that plaintiffs have "declared [their] intention to take no further action,” ante at 1157, this has little to do with the facts and holding in Borelli v. City of Reading, 532 F.2d 950, 951-52 (3d Cir.1976), on which the majority relies. In Borelli, the district court dismissed the complaint without prejudice because the plaintiff had failed to establish standing. Although this Court noted in dictum that "if the plaintiff cannot amend or declares his intention to stand on his complaint,” then the order “becomes final and appealable,” we dismissed the appeal for want of jurisdiction, because “[w]e needfed to] go no further than to note that we [were] not confronted with a situation where it [was] not possible," id. at 952 (emphasis added), for the plaintiff “to take further action."
. This case itself serves as an illustration of that point. The district court’s non-final order was entered on September 18, 1985. Thus, more than a full year has been lost from the plaintiffs’ taking of this interlocutory appeal. Had the parties proceeded to a final disposition, and then appealed the issues in question, those issues would have been resolved long ago.