(dissenting). I would reverse the judgment and order a new trial because the court’s reasonable doubt charge incorporated language that diluted the People’s burden of proof and, even when read in conjunction with the remainder of the charge, created a substantial risk of confusion.

We are unanimous in the conclusion that the portion of the reasonable doubt charge of which defendant complains was improper. As is implicit in the majority’s holding, the disputed instruction conveys the false impression that the People may prevail if the evidence weighs slightly more heavily in their[*824] favor (see, e.g., People v Weatherspoon, 155 AD2d 888; People v Hartle, 151 AD2d 1003; People v Celestin, 150 AD2d 385). Further, the use of the term "innocent” as a substitute for the term "not guilty” is offensive because it suggests that the defense must make some showing and cannot rely solely on the presumption of innocence.

Despite this serious flaw in the court’s charge, the majority concludes that reversal is not required because the charge "as a whole” conveyed the correct standard of proof. In my view, that conclusion does not withstand analysis on this record.

It is true as the majority notes that a reviewing court should not consider a single instruction "alone and in a vacuum” (majority mem, at 823), but that observation does not, standing by itself, resolve the specific issue presented here. No instruction dealing with the "reasonable doubt” standard is ever issued "in a vacuum”; such instructions — whether accurate or misleading — are almost always issued as part of a series of sentences describing what constitutes reasonable doubt. In most cases in which a portion of the reasonable doubt charge is challenged, the trial court has given the boilerplate language contained in the Pattern Jury Instructions and then has added some improvised language of its own. The fact that the standard, previously approved reasonable doubt instruction has also been given cannot alone be sufficient to neutralize an otherwise improper or misleading instruction on the concept. To the contrary, each situation should be examined on its own facts to ascertain whether the erroneous instruction created a risk that the jury would be misled.

In making this assessment, the reviewing court must be sensitive to the fact that reasonable doubt is "a nebulous concept not susceptible of precise definition” (People v Antommarchi, 80 NY2d 247, 251). Consequently, it is necessarily explained in general or abstract terms that some may find difficult to apply (id.). Furthermore, because the standard plays such a vital and, indeed, "indispensable” role in our system of criminal justice (see, In re Winship, 397 US 358, 363), reviewing courts should scrutinize a potentially misleading reasonable doubt charge with particular care.

Here, the trial court first instructed the jury on the general concept of reasonable doubt, defining it as a doubt arising from the evidence or lack thereof for which some reason can be given. It then stated, in somewhat circular fashion, that "[t]he doubt must be one that a reasonable person, acting in a matter such as this, would be likely to have.” The court then went on[*825] to enumerate what is not a "reasonable doubt” and explained that "[i]f the People have not proved the [djefendant’s guilt beyond a reasonable doubt, then you should find him not guilty.” The presumption of innocence was correctly charged, and the court reiterated the People’s burden of proving the defendant’s guilt beyond a reasonable doubt. The very last point that the jury heard on the reasonable doubt standard was the instruction which, as we all agree, was improper.

Since the offending instruction came at the end of the reasonable doubt charge, there is a very real danger that the jurors regarded it as the "last word” and the most definitive explanation of the concept. Moreover, there was a substantial risk that the erroneous charge had a disproportionate influence on the jurors’ understanding of reasonable doubt, since it was more vivid than much of the hazy and abstract definition that the court gave. Significantly, most of the remaining charge on reasonable doubt explained the concept by using one or more of the words "reason”, "reasonable” or "doubt” and by explaining what a "reasonable doubt” is not. A juror struggling with the "nebulous” concept of reasonable doubt might well have grasped onto the court’s statement about evenly divided evidence as the most readily understandable explanation of the concept — and the only one that was not self-referential — and might then have applied it without further thought to the other, murkier portions of the court’s reasonable doubt charge.

Significantly, the unobjectionable part of the trial court’s reasonable doubt charge contained no language which contradicted the impermissible inference implicit in the objected-to charge, i.e., that if the preponderance of the evidence favors the People, a guilty verdict is the proper result. Thus, there was no reason for the jurors even to suspect that the foregoing is an erroneous standard (cf., People v Warren, 76 NY2d 773, 775).

For all of the foregoing reasons, I disagree with the majority’s conclusion that the charge here unequivocally conveyed the proper legal standard and that there is no serious risk that the jury was misled by the burden-diluting part of the court’s reasonable doubt charge. Because an error in the reasonable doubt charge vitiates the reliability of the verdict as a whole (see, Sullivan v Louisiana, 508 US 275, 278-281), I would reverse the judgment of conviction and order a new trial.

[*826] Chief Judge Kaye and Judges Simons, Bellacosa, Smith, Levine and Ciparick concur; Judge Titone dissents and votes to reverse in an opinion.

Order affirmed in a memorandum.