An adapted alternating treatments design was used to evaluate and compare the effects of two group contingency interventions on mathematics assignment accuracy in an intact first-grade classroom. Both an interdependent contingency with class-average criteria (16 students) and a dependent contingency with criteria based on the average of a smaller, unknown, randomly selected group of students (4 students) were applied. For both contingencies, rewards and criteria were randomly selected and unknown to students. Results showed immediate, sustained, and meaningful improvements in mathematics assignment accuracy (from a class average of 64% to a class average above 83%) across both contingencies, with little differences between the two interventions. Social validity data suggest that the two teachers and the majority of the students preferred the small-group contingency. Discussion focuses on applied implications of the current results and directions for future research, including investigating side effects and idiosyncratic effects.