The social interaction between a behaviorally disordered preschool child and her classroom peers was measured under two conditions of contingent adult attention: (1) verbal praise and physical contact directed to the target subject's peers for appropriate interaction with the target subject, and (2) verbal praise and physical contact directed specifically to the target subject for engaging in appropriate interaction with peers. Continuous measures of interactive behavior were made during baseline, intervention, and return to baseline conditions. Results indicated that application of experimental contingencies to peers (Condition 1) rapidly increased appropriate social behaviors by the peers and also by the target subject. When experimental contingencies were applied to the target subject (Condition 2), a similar increase in appropriate social behaviors was noted for both the target subject and the peers. Additionally, during Conditions 1 and 2 the recipient(s) of contingent adult attention initiated more appropriate social contacts than did the interacting partner(s).