with whom Justice Stevens, Justice Ginsburg, and Justice Breyer join, dissenting.
The majority holds that a testifying defendant perforce waives the right to appeal an adverse in limine ruling admitting prior convictions for impeachment. The holding is without support in precedent, the rules of evidence, or the reasonable objectives of trial, and I respectfully dissent.
The only case of this Court that the majority claims as even tangential support for its waiver rule is Luce v. United States, 469 U. S. 38 (1984). Ante, at 759. We held there that a criminal defendant who remained off the stand could. not appeal an in limine ruling to admit prior convictions as impeachment evidence under Federal Rule of Evidence 609(a). Since the defendant had not testified, he had never suffered the impeachment, and the question was whether he should be allowed to appeal the in limine ruling anyway, on the rationale that the threatened impeachment had discouraged the exercise of his right to defend by his own testimony. The answer turned on the practical realities of appellate review.
An appellate court can neither determine why a defendant refused to testify, nor compare the actual trial with the one that would have occurred if the accused had taken the stand. With unavoidable uncertainty about whether and how much the in limine ruling harmed the defendant, and whether it affected the trial at all, a rule allowing a silent defend*761ant to appeal would require courts either to attempt wholly speculative harmless-error analysis, or to grant new trials to some defendants who were not harmed by the ruling, and to some who never even intended to testify. In requiring testimony and actual impeachment before a defendant could appeal an in limine ruling to admit prior convictions, therefore, Luce did not derive a waiver rule from some general notion of fairness; it merely acknowledged the incapacity of an appellate court to assess the significance of the ruling for a defendant who remains silent.
This case is different, there being a factual record on which Ohler’s claim can be reviewed. She testified, and there is no question that the in limine ruling controlled her counsel’s decision to enquire about the earlier conviction; defense lawyers do not set out to impeach their own witnesses, much less their clients. Since analysis for harmless error is made no more difficult by the fact that the convictions came out on direct examination, not cross-examination, the case raises none of the practical difficulties on which Luce turned, and Luce does not dictate today’s result.1
In fact, the majority’s principal reliance is not on precedent but on the “commonsense” rule that “a party introducing evidence cannot complain on appeal that the evi*762dence was erroneously admitted.” Ante, at 755. But this is no more support for today’s holding than Luce is, for the common sense that approves the rule also limits its reach to a point well short of this case. The general rule makes sense, first, when a party who has freely chosen to introduce evidence of a particular fact later sees his opponent’s evidence of the same fact erroneously admitted. He suffers no prejudice. See Mercer v. Theriot, 377 U. S. 152, 154 (1964) (per curiam); 21 C. Wright & K. Graham, Federal Practice and Procedure § 5039, p. 203 (1977). The rule makes sense, second, when the objecting party takes inconsistent positions, first requesting admission and then assigning error to the admission of precisely the same evidence at his opponent’s behest. “The party should not be permitted ‘to blow hot and cold’ in this way.” 1 J. Strong, McCormick on Evidence § 55, p. 246, n. 14 (5th ed. 1999).
Neither of these reasons applies when (as here) the defendant has opposed admission of the evidence and introduced it herself only to mitigate its effect in the hands of her adversary. Such a case falls beyond the scope of the general principle, and the scholarship almost uniformly treats it as exceptional. See, e. g., 1 J. Wigmore, Evidence § 18, p. 836 (P. Tillers rev. 1983) (“[A] party who has made an unsuccessful motion in limine to exclude evidence that he expects the proponent to offer may be able to first offer that same evidence without waiving his claim of error”); M. Graham, Handbook of Federal Evidence § 103.4, p. 17 (1981) (“However, the party may . . . himself bring out evidence ruled admissible over his objection to minimize its effect without it constituting a waiver of his objection”); 1 McCormick, supra, §55, at 246 (“[W]hen [a party’s] objection is made and overruled, he is entitled to treat this ruling as the ‘law of the trial’ and to explain or rebut, if he can, the evidence admitted over his protest”); D. Louisell & C. Mueller, Federal Evidence §11, p. 65 (1977) (“Having done his best by objecting, the adversary would be indeed ill treated if then *763he was held to have thrown it all away by doing his best to protect his position by offering evidence of his own”).2 The general thrust of the law of evidence, then, not only fails to support the majority’s approach, but points rather clearly in the other direction.
With neither precedent nor principle to support its chosen rule, the majority is reduced to saying that “there is nothing ‘unfair’ . . . about putting petitioner to her choice in accordance with the normal rules of trial.”3 Ante, at 759. Things are not this simple, however.
Any claim of a new rule’s fairness under normal trial conditions will have to stand or fall on how well the rule would serve the objects that trials in general, and the Rules of Evidence in particular, are designed to achieve. Thus the provisions of Federal Rule of Evidence 102, that “[tjhese rules shall be construed to secure fairness in administration, elimination of unjustifiable expense and delay, and promotion of growth and development of the law of evidence to the end that the truth may be ascertained and proceedings justly determined.” A judge’s job, accordingly, is to curb the tactics of the trial battle in favor of weighing evidence calmly and getting to the most sensible understanding of whatever gave rise to the controversy before the court. The question is not which side gains a tactical advantage, but which rule assists in uncovering the truth. Today’s new rule can make no such claim.
*764Previously convicted witnesses may testify honestly, but some convictions raise more than the ordinary question about the witness’s readiness to speak truthfully. A fact-finder who appreciates a heightened possibility of perjury will respond with heightened scrutiny, and when a defendant discloses prior convictions at the outset of her testimony, the jury will bear those convictions in mind as she testifies, and will scrutinize what she says more carefully. The purpose of Rule 609, in making some convictions admissible to impeach a witness’s credibility, is thus fully served by a defendant’s own testimony that the convictions occurred.
It is true that when convictions are revealed only on cross-examination, the revelation also warns the factfinder, but the timing of their disclosure may do more. The jury may feel that in testifying without saying anything about the convictions the defendant has meant to conceal them. The jury’s assessment of the defendant’s testimony may be affected not only by knowing that she has committed crimes in the past, but by blaming her for not being forthcoming when she seemingly could have been. Creating such an impression of current deceit by concealment is very much at odds with any purpose behind Rule 609, being obviously antithetical to dispassionate factfinding in support of a sound conclusion. The chance to create that impression is a tactical advantage for the Government, but only in the majority’s dismissive sense of the term; it may affect the outcome of the trial, but only if it disserves the search for truth.
Allowing the defendant to introduce the convictions on direct examination thus tends to promote fairness of trial without depriving the Government of anything to which it is entitled. There is no reason to discourage the defendant from introducing the conviction herself, as the majority’s waiver rule necessarily does.
The Luce Court anticipated as much: “It is clear, of course, that had petitioner testified and been impeached by evidence of a prior conviction, the District Court’s decision to admit the impeachment evidence would have been reviewable on appeal along with any other claims of error. The Court of Appeals would then have had a complete record detailing the nature of petitioner’s testimony, the scope of the cross-examination, and the possible impact of the impeachment on the jury’s verdict.” 469 U. S., at 41. There are, of course, practical issues that may arise in these cases; for example, the trial court may feel unable to render a final and definitive in limine ruling. The majority does not focus on these potential difficulties, and neither do I, though some lower courts have addressed them. See, e. g., Wilson v. Williams, 182 F. 3d 562 (CA7 1999) (en bane). For the purposes of this case, we need consider only the circumstance in which a district court makes a ruling that is plainly final.
The point on which the analysis of the cited treatises turns, it should be clear, is not which party first introduces the evidence, but rather which party seeks introduction and which exclusion. A defense lawyer who elicits testimony about prior convictions on direct examination, having failed in an attempt to have them excluded, is plainly making a defensive use of the convictions; he has no desire to impeach his client. The fact that it is the defense lawyer who first introduces the convictions, then, is irrelevant to the principle the majority invokes.
For the reasons just given, this begs the question, which is whether the “normal rules of trial” apply beyond the normal circumstances for which they were devised.