The current study evaluated the effects of an interdependent group contingency with randomized components on student behavior during the transition from lunch to class. The study was conducted in three high school classrooms in an alternative school setting for students with emotional and behavioral disorders and used an ABAB withdrawal design. During the intervention, the teacher informed the students that it was time for class to begin and then scanned the room to count the number of students ready (i.e., seated at assigned desk, eyes directed at the teacher or front of the room, and not being physically or verbally disruptive). The teacher randomly selected a criterion number of students who needed to be ready to start instruction. If the class met the criterion, the teacher randomly selected a reward. Dependent variables included the percentage of students ready to begin instruction within 5 min of the official class start time and the number of minutes past the official start time that class actually began (i.e., teacher directed students to an academic task or explicitly stated that class was starting). Across all classrooms, results indicated significant improvements in student transition behavior. Discussion focuses on the practical implications for providing intervention supports to students with problem behavior.