720 college students served on 6- or 12-member "juries" that listened to a simulated trial of a rape case, deliberated, and then rendered verdicts according to the unanimity or two-thirds majority social decision rule assigned. Although neither assigned rule nor jury size had a significant effect upon the verdict distribution, average deliberation time and the number of polls were both significantly influenced. Individuals gave a significantly higher proportion of guilty verdicts than did juries. A two-thirds majority social decision scheme model best predicted overall verdict distributions, while a number of other social process models could be confidently rejected. Several other findings are discussed along with the argument for the need for both theoretical and empirical research on juries. (28 ref) (PsycINFO Database Record (c) 2016 APA, all rights reserved)