We evaluate an intervention designed to increase teachers' awareness of social isolation by providing them with their own students' social network and information on developmental risks associated with social exclusion. Using friendship data and incentive-compatible measures of antisocial and prosocial behavior, we find that the intervention reduces social isolation and antisocial behavior without improving prosocial behavior. The reduction in antisocial behavior leads to better economic outcomes in treated classrooms, measured by average payoffs and the Gini coefficient. Our findings highlight the personal and communal benefits of alleviating social exclusion and antisocial peer relationships in schools.