BOTH EXPERIMENTS INVOLVED A SIMULATED TRIAL FOR AGGRAVATED AND FORCIBLE RAPE. THE VICTIM'S AND DEFENDANT'S RACE (BLACK OR WHITE) AND THE AMOUNT OF EVIDENCE POINTING TOWARD GUILT WERE SYSTEMATICALLY VARIED. THE SUBJECTS OF THE FIRST EXPERIMENT WERE 244 WHITE UNDERGRADUATES AT A MIDWESTERN UNIVERSITY. IN THE SECOND EXPERIMENT, THE SUBJECTS WERE 186 BLACK UNDERGRADUATES. THE AMOUNT OF EVIDENCE, THE SEX OF THE SUBJECT, THE RACE OF THE VICTIM, AND THE RACE OF THE DEFENDANT ALL INFLUENCED THE LEVEL OF CULPABILITY ASCRIBED TO THE DEFENDANT BY BOTH WHITE AND BLACK JURORS. WHEN THERE WAS ONLY MARGINAL EVIDENCE AGAINST THE DEFENDANT, JURORS WERE MORE LIKELY TO ATTRIBUTE GUILT TO A DEFENDANT RACIALLY DIFFERENT FROM THEMSELVES THAN TO A DEFENDANT OF THEIR OWN RACE. IN THE STRONG EVIDENCE CONDITION, BLACKS BUT NOT WHITES SHOWED A SIGNIFICANT BIAS IN FAVOR OF OWN-RACE DEFENDANTS. THE FINDINGS SUPPORT THE IDEA THAT AMBIGUITY IN THE FACTS OF A CASE SERVES TO FREE JURORS TO RESPOND TO THEIR RACIAL PREJUDICES AND BIASES. THE TENDENCY FOR BLACK SUBJECTS TO MINIMIZE THE CULPABILITY OF BLACK DEFENDANTS EVEN IN THE PRESENCE OF STRONG INCRIMINATING EVIDENCE MAY REFLECT THE SUBJECTS' OWN EXPERIENCES WITH THE LEGAL SYSTEM AND OR THEIR IDEOLOGIES. SEVERAL SOCIAL AND PSYCHOLOGICAL FACTORS PROBABLY ACCOUNT FOR THE FINDINGS THAT FEMALES RESPONDED MORE PUNITIVELY THAN MALES, AND THAT WHITE JURORS RESPONDED MOST PUNITIVELY IN CASES INVOLVING BLACK DEFENDANTS AND WHITE VICTIMS. SUPPORTING DATA AND A LIST OF REFERENCES ARE INCLUDED. (LKM)