Students may engage in behavior during instruction that impacts the acquisition of skills. Intervening on student behavior using group contingencies may be more efficacious than other intervention methods. Directly programming for generalization of treatment effects may increase the durability of behavior change and further increase the efficacy of group contingencies. The current study examined the effect of a randomized dependent group contingency and programming common stimuli on levels of academic engagement and problem behavior for second-grade participants receiving small-group reading and writing instruction. The results showed that a randomized dependent group contingency increased the academic engagement of primary participants and several of the other participants during small-group instruction. The findings also showed that high levels of academic engagement were maintained when common stimuli were present and the dependent group contingency was withdrawn.