This study examined the social relationships and behavior of physically abused schoolchildren. Its emphasis on peer relationships was based on the fact that abused children’s basic socializing and support system -- their relationship with family -- was expected to be damaged and give rise to internalizing and externalizing problems. Their peer networks therefore were expected to play a disproportionate role in their adaptive functioning in many domains of development. Family relationships, operating through social learning and social cognitive processes, were expected to influence children’s social behavior, giving rise to aggressive and antisocial behavior. Such behavior was hypothesized to raise the risk for lowered social status with peers which in turn was expected to lead to internalizing and externalizing problem behaviors. For abused children who managed to establish good relationships with peers, such relationships might mitigate the effects of abuse on later functioning. The sample consisted of 100 physically abused urban schoolchildren (65 boys, 35 girls) ages 9-12 years and in grades 4-6, and 100 non-abused classmates case-matched for gender, age, and, as closely as possible, for race, ethnicity, and socioeconomic status. Control subjects were screened for abuse by interviews with their caretakers about the handling of disputes among household members, and by scanning the Abuse Register to ascertain that their names did not appear during the 4 years we were recruiting abuse cases. Abused children were recruited from confirmed cases of physical abuse in consecutive entries onto the Abuse Register from 1992 to 1996. Children who were sexually abused were excluded, but children who were neglected as well as physically abused were not. The first 100 families meeting study criteria and agreeing to participate were enrolled. Practically all the children were of minority status. Sociometric assessments were carried out in the 100 classrooms of the abuse/control pairs to determine subjects’ sociometric status among same-gender classmates; interviews were conducted with the children about their friends and understanding of social relationships; interviews were conducted with their parents about family and household demographics, family stressful life events, mental health of the child’s major caretaker, and intra-family relationships; classmates rated the children’s prosocial and antisocial behavior; and teachers and parents rated children’s problem behavior. Results indicated that children’s social expectations regarding peers, and two social behaviors -- aggressive and prosocial behavior -- mediated between abuse and positive and negative social status, and between abuse and positive and negative reciprocity. Social expectations and withdrawn behavior mediated between abuse and positive social status. Social expectations and negative social status (peer rejection) mediated between abuse and internalizing problems. Acknowledging that family contextual factors are important influences on child outcome, we proposed an ecological model that designated family stress as the principal exogenous factor, with effects on outcome mediated through caretaker distress, partner violence, and physical child abuse. Outcomes included parent-, teacher- and peer-rated child behavior. Results were consistent with the hypothesis that partner violence and caretaker distress, both associated with family stress, increase the risk for child abuse and thereby raise the child’s risk for problem behaviors. Investigators: Suzanne Salzinger, Ph.D. New York State Psychiatric Institute New York, NY Richard Feldman, Ph.D. New York State Psychiatric Institute New York, NY Daisy S. Ng-Mak, Ph.D. Columbia University New York, NY