THE EXTERNAL VALIDITY OF EXPERIMENTS TO DETERMINE HOW JURIES MAKE DECISIONS CONCERNING THE GUILT OR INNOCENCE OF DEFENDANTS IS PARTICULARLY IMPORTANT SINCE THE FINDINGS HAVE POTENTIALLY SERIOUS CONSEQUENCES. THE PRESENTATION AND CONTENT OF EXPERIMENTAL MANIPULATIONS OR SIMULATIONS DIFFER MARKEDLY FROM ACTUAL TRIALS BECAUSE THE NATURE OF THE STIMULUS PRESENTED TO THE SUBJECTS IS USUALLY IN THE FORM OF A SHORT WRITTEN DESCRIPTION INCLUDING NO EXAMINATION AND CROSS-EXAMINATION OF WITNESSES AND NO JUDICIAL INSTRUCTIONS. THE MOST SERIOUS PROBLEM WITH INDEPENDENT VARIABLES IN JURY DECISIONMAKING RESEARCH, HOWEVER, IS CONCEPTUAL RATHER THAN METHODOLOGICAL; THE IMPORTANT QUESTION IS WHETHER JURORS FIND IRRELEVANT CRITERIA SIGNIFICANT EVEN WHEN RELEVANT CRITERIA ARE AVAILABLE FOR DETERMINING A VERDICT. OTHER DIFFICULTIES PERTAINING TO THE EXTERNAL VALIDITY OF THESE EXPERIMENTS INVOLVE THE USE OF STUDENTS AS SUBJECTS AND THE FACT THAT EXPERIMENTAL JURORS ARE NOT ALLOWED TO DELIBERATE. AN EXPERIMENT WAS DESIGNED TO ASSESS THE RELATIVE IMPORTANCE OF LEGALLY RELEVANT AND IRRELEVANT CRITERIA IN JURY DECISIONMAKING. GUILT WAS THE DEPENDENT VARIABLE, AND THE GUILT OF THE DEFENDANT WAS CONSIDERED TO BE AMBIGUOUS. THE TRIAL WAS PRESENTED TO EXPERIMENTAL SUBJECTS BY MEANS OF A VIDEOTAPE CONSISTING OF WITNESS TESTIMONY, CROSS-EXAMINATION TESTIMONY, AND INSTRUCTIONS OF THE JUDGE. THE VIDEOTAPE WAS PRECEDED BY A SHORT ORAL DESCRIPTION OF CASE FACTS AND THE TASK OF THE JURY. THREE INDEPENDENT VARIABLES (INTENT, INSTRUCTION, AND VICTIM) WERE CROSSED IN THE EXPERIMENTAL DESIGN TO PRODUCE EIGHT VERSIONS OF THE TRIAL. THESE EIGHT VERSIONS WERE PRESENTED TO FOUR MAJOR GROUPS OF SUBJECTS: DELIBERATING JURORS, DELIBERATING STUDENTS, NONDELIBERATING JURORS, AND NONDELIBERATING STUDENTS. IT WAS FOUND THAT LEGALLY RELEVANT CRITERIA WERE MOST IMPORTANT IN PRODUCING JURY VERDICTS. THIS EFFECT WAS MOST EVIDENT IN THE DECISIONS OF THE DELIBERATING JURORS AND TO A LESSER EXTENT IN THE DECISIONS OF THE DELIBERATING STUDENTS. STUDENTS APPEARED TO CONSIDER BOTH THE LEGALLY RELEVANT AND IRRELEVANT CRITERIA TO BE IMPORTANT IN DETERMINING VERDICTS. METHODOLOGY APPEARED TO BE IMPORTANT IN DETERMINING VERDICTS AND IN THE DECISIONMAKING PROCESS LEADING TO VERDICTS. DELIBERATING JURORS WERE THE MOST LIKELY TO CONFORM TO THE LEGAL MODEL OF DECISIONMAKING. SUPPORTING DATA ON THE EXPERIMENTAL STUDY, A LIST OF REFERENCES AND FOOTNOTES ARE INCLUDED. (DEP)