Appeal by defendant from a judgment of the Supreme Court, Suffolk County (Mclnerney, J.), rendered June 23,1983, as amended June 29, 1983, convicting him of grand larceny in the third degree, petit larceny, and issuing a bad check (two counts), upon a jury verdict, and imposing sentence. *916f Judgment as amended reversed, on the law and as a matter of discretion in the interest of justice, and new trial ordered. 11 Defendant was tried on charges stemming from incidents on October 11, 1981 and November 9, 1981 when defendant allegedly issued two bad checks in the amount of $81.61 and $411.65, respectively, in violation of section 190.05 of the Penal Law. At trial, as part of their proof to establish necessary intent, the People were permitted, over defense objections, to show the entire history of defendant’s checking account and to prove that some 15 other checks were dishonored during the same approximate time period as the two in question. Defendant’s principal contention on appeal is that Criminal Term erred in permitting the People to bolster their case through the introduction of this evidence. We agree, f It is fundamental that evidence of uncharged crimes is not admissible at trial if the sole purpose is to show the defendant’s predisposition to commit the crime charged (see People v Allweiss, 48 NY2d 40; People v Molineux, 168 NY 264). On the other hand, evidence relevant to prove some fact in the case other than propensity, e.g., intent, is not inadmissible merely because it may also reveal that the defendant has committed other crimes (see People v Fiore, 34 NY2d 81; People v Molineux, supra). Before permitting such evidence of uncharged crimes, however, the trial court must make a determination that the probative value of the evidence outweighs its potential prejudice to the defendant (see People v McKinney, 24 NY2d 180), preferably at a hearing held in limine where the court can “assess how the evidence came into the case and the relevance and probativeness of, and necessity for it against its prejudicial effect, and either admit or exclude it in total, or admit it without the prejudicial parts when that can be done without distortion of its meaning” (People v Ventimiglia, 52 NY2d 350, 362; see, also, People u Santarelli, 49 NY2d 241, 249-250). 11 Turning to the case at bar, we note initially that although the record refers to some “discussion in chambers on People versus Molineux, and certain preliminary rulings * * * made”, it fails to contain any hearing where the court weighed the relevance and probativeness of the evidence as against its prejudicial impact. I Even assuming, arguendo, that such a hearing was duly held, we are of the opinion that Criminal Term abused its discretion in letting in the evidence of these 15-odd checks, not subject of any count in the indictment. We note that this evidence was not introduced to show the existence of a common scheme or plan. Further, testimony concerning a lesser number of checks might well have been permissible to establish defendant’s intent (see People v Molineux, supra, pp 297-298). But the wholesale introduction of testimony relating to 15 dishonored checks could only have overwhelmed the jury and created in its mind the impression of defendant’s predisposition to issue bad checks (see People vZeldes, 78 AD2d 865). The resulting prejudice to defendant was made virtually total by the fact that Criminal Term permitted the testimony in question without even instructing the jury as to the limited purpose for which it was being offered. Only later, during the formal jury charge, and then too little, too late, did the court offer any limiting instruction as to this testimony. On this record, we simply cannot be sure that the jury did not “convict to punish the person portrayed by the evidence before them even though [they were] not convinced beyond a reasonable doubt of his guilt of the crime[s] * * * charged” (People v Ventimiglia, supra, p 359). 1i Accordingly, the judgment of conviction must be reversed and a new trial ordered. Titone, J. P., O’Connor, Brown and Eiber, JJ., concur.