Background: Violence exposure within each setting of community, school, or home has been linked with internalizing and externalizing problems. Although many children experience violence in multiple contexts, the effects of such cross-contextual exposure have not been studied. This study addresses this gap by examining independent and interactive effects of witnessing violence and victimization in the community, home, and school on subsequent internalizing and externalizing problems in early adolescence. Methods: A community sample of 603 boys and girls (78% African American, 20% Caucasian) participated in a longitudinal study of youth violence. During two assessments 16 months apart, adolescents reported on witnessing violence and victimization in the community, school, and home, and their internalizing and externalizing problems. Results: Multiple regressions tested the independent and interactive effects of witnessing violence or victimization across contexts on subsequent adjustment, after controlling for initial levels of internalizing and externalizing problems and demographic covariates. Witnessing violence at school predicted anxiety and depression; witnessing at home was related to anxiety and aggression; and witnessing community violence predicted delinquency. Victimization at home was related to subsequent anxiety, depression, and aggression; victimization at school predicted anxiety; and victimization in the community was not independently related to any outcomes. Finally, witnessing violence at home was associated with more anxiety, delinquency, and aggression only if adolescents reported no exposure to community violence. Conclusions: Violence exposure at home and school had the strongest independent effects on internalizing and externalizing outcomes. Witnessing community violence attenuated the effects of witnessing home violence on anxiety and externalizing problems, perhaps due to desensitization or different norms or expectations regarding violence. However, no comparable attenuation effects were observed for victimization across contexts.