The purpose of this study was to investigate the effects of blocked and random practice schedules on the performance accuracy, speed, temporal evenness, and attitude of beginning band students in a group instructional setting. The research assumptions were based on the contextual interference hypothesis, which predicts that a blocked practice order (low contextual interference) leads to superior performance immediately following practice but that a random practice order (high contextual interference) supports superior performance at delayed retention testing. Beginning clarinet students (N = 41) completed three practice sessions and one retention testing session, performing three seven-pitch exercises. At the end of practice, no significant differences were found between blocked and random practice groups for accuracy, speed, or temporal evenness. At retention, the random group performed significantly faster than the blocked group, F(1, 38) = 24.953, p less than 0.001, [eta][superscript 2] = 0.92, and the blocked group performed significantly slower than it did at the end of practice (p less than 0.001). No significant differences were found between groups for transfer tasks or for attitude toward practice. (Contains 2 figures, 1 table and 1 note.)