EACH OF THE 214 COLLEGE STUDENTS PARTICIPATING IN THE STUDY OF 1 OF 8 VERSIONS OF A TRIAL INVOLVING 3 DEFENDANTS CHARGED WITH BREAKING AND ENTERING AND POSSESSION OF MARIJUANA. VARIABLES WERE THE LEVEL OF INCRIMINATING EVIDENCE, AND THE PRESENCE OR ABSENCE OF JUDGE'S INSTRUCTIONS AND ATTEMPTS BY THE PROSECUTING ATTORNEY TO DISCREDIT DEFENSE WITNESSES. JUDGE'S INSTRUCTIONS AND LEVEL OF INCRIMINATING EVIDENCE, BUT NOT PROSECUTOR'S IMPEACHMENT STRATEGY, EXERTED A SIGNIFICANT EFFECT ON VERDICTS. JURORS WITHOUT INSTRUCTIONS WERE MORE LIKELY TO VOTE GUILTY OR REPORT NO DECISION. JUDGE'S INSTRUCTIONS APPEARED TO MOVE SUBJECTS FROM INDECISION OR CONVICTION TO MORE LENIENT VOTING. IN ADDITION, 70 PERCENT OF INSTRUCTED SUBJECTS WHO RECORDED NO DECISION CHANGED TO A VERDICT OF GUILTY OR NOT GUILTY WHEN ADDITIONAL INSTRUCTIONS WERE GIVEN. THE STUDY FINDINGS SUGGEST THAT JUDICIAL INSTRUCTIONS SHOULD BE CONSIDERED IN FURTHER RESEARCH ON JURY DECISIONMAKING. ONE INCIDENTAL FINDING WAS THE RELATIVELY LARGE NUMBER OF COLLEGE STUDENTS VOTING GUILTY IN A CASE INVOLVING MARIJUANA CHARGES--A RESULT THAT BRINGS INTO QUESTION PROSECUTORS' AUTOMATIC REJECTION OF YOUNG PERSONS AS JURORS IN DRUG CASES. SUPPORTING DATA AND FOOTNOTES ARE INCLUDED.