Results indicate that presenting a legal definition of grand larceny before, after, and before and after testimony influenced decisionmaking, resulting in fewer guilty verdicts relative to the no-definition condition. Before and after definitions resulted in significantly fewer guilty verdicts (12.5 percent) than did presenting them before (27.5 percent) or after (30 percent). No significant differences in memory for relevant legal testimony were found among groups, suggesting results were due to information used to decide guilt rather than selective encoding of information. All subjects who returned a guilty verdict justified the verdict on the basis of the defendant's action, while those returning not-guilty verdicts used either the legal definition or moral arguments to justify their decisions. 12 references.