In this study, three variations of a withdrawal design were used to assess the effects that group instruction, peer-dyads, and peer-mediated accountability had on the number of trials performed, and how successful those trials were, during one-minute trials of volleyball skills. Peer-mediated accountability consisted of teacher-established goals, peer recording of performance, public posting of student performance, and special content-related activities that served as public recognition of achievement. Participants were 67 elementary school students in grades 4 through 6. Results indicated that students performed more trials and were generally more successful in the peer-mediated accountability condition than during either the peer-dyads or group instruction. Findings are discussed in terms of the contingent relation between tasks and consequences created by the peer-mediated accountability variable.