This study evaluated the effectiveness of Resilient Peer Treatment (RPT). This is a peer-mediated, classroom-based intervention for socially withdrawn, maltreated preschool children. It examined whether the RPT impact generalized from the treatment setting to larger classroom context. Eighty-two maltreated and nonmaltreated, socially withdrawn Head Start children were randomly assigned to either RPT or attention-control (AC) conditions. Data were collected by teachers and independent observers blind to both maltreatment status and treatment condition. Treatment resulted in higher levels of collaborative peer play interactions in the treatment setting posttreatment for both the maltreated and nonmaltreated children. Results documented generalization of the treatment impact to classroom free-play sessions. These findings were supported by teacher ratings of interactive peer play and social skills.