Group contingencies are an effective way to prevent and address problem behavior; however, the extent to which these effects transfer to settings where the intervention is not implemented (i.e., setting generalization) is unknown. Using a multiple baseline design, this study assessed the setting generalization of the effects of an interdependent group contingency on decreasing off-task behavior in elementary students with emotional and behavioral disorders. A combination of generalization strategies was used, including maximizing the response's contact with reinforcement, making the instructional setting similar to the generalization setting, training multiple exemplars, and training to generalize. Participants included three classrooms (N = 22 students; 3 female, 19 male) in grades K--6. Results showed that intervention effects did not automatically generalize across settings. With the addition of the generalization strategies, behavior in the generalization setting reached levels similar to the instructional setting, indicating the strategies were successful. Implications for practice and research are discussed.