Due to the passage of Public Law 94-142 (1975), widespread attempts have been made to integrate children who have handicaps into settings with their nonhandicapped peers. Although integrated settings may provide the opportunity for social interaction to occur between children with and without handicaps, often interaction does not occur. In order to address the issue of how to best facilitate appropriate interactions in integrated settings, specialized programs such as the FMS buddy system (Quintero, Phelps, Striefel, & Killoran, 1987) have been developed to promote positive social interactions between children with and without handicaps. One important aspect programs such as the buddy system have not fully considered is the differential impact a nonhandicapped child could have on the level of social interaction of the child with handicaps. The impact the nonhandicapped buddy could make if constantly paired with the same child with handicaps may be different than the one a nonhandicapped child could make if paired, over time, in an alternating sequence with different children who have handicaps. In response to the question of possible differential impact, a single subject multiple baseline design was utilized to compare the effect constant buddies and rotating buddies had on the social interactions of 8 children who had handicaps. The intervention included training the buddies on how to interact with children who have handicaps and providing the opportunity for the children with and without handicaps to play together. Treatment effects were measured by direct observations of social interactions between the children with and without handicaps during free play and buddy sessions, sociometric measures, and attitude measures. Results indicated that pairing children who have handicaps with a nonhandicapped buddy increased social interactions between children with and without handicaps during buddy sessions. The level of interaction achieved during buddy sessions was not fully generalized to subsequent free play sessions. Buddies from constant dyads rated their playmates who had handicaps sociometrically higher than buddies from rotating dyads. Non handicapped children who served as buddies rated their buddy experience favorably. Suggestions for future research in this area are made.