Multivariate analyses of self-reports and teacher reports of various types of aggressive behavior showed both mediating and independent effects of social information processing and experiences of aggression in social contexts. Having an aggressive-impulsive response repertoire for behaviors in various social interactions was linked to self-reports and teacher reports of physical aggression, verbal aggression, violent offenses, general aggression, and other forms of delinquency. Positive evaluations of aggressive responses showed a weaker effect on all types of self-reported and teacher-reported aggression. Attributions of hostility and aggressive/egocentric goal-setting had no impact on reports of aggression. Perceived aggression in the family, in the peer group, in media consumption, and at school predicted verbal aggression, physical aggression, and violent offenses. These findings are discussed from methodological and theoretical perspectives. The study involved 102 male adolescents who were selected for a longitudinal and more detailed analysis from a representative sample of 1,163 students attending schools at Nuremberg and Erlangen, Germany. At Time 1 (grades seven or eight), the boys answered questions related to three written scenarios that involved conflicts which could trigger aggressive responses from individuals with aggression-prone attitudes/perceptions. The boys also reported on the aggression and conflict they experienced in various social contexts (family, school, peer group, and media consumption). At both Time 1 and Time 2 (grades 9 and 10), self-reports and teacher reports were obtained on the boys' physical aggression, verbal aggression, and delinquent behavior. 5 tables and 86 references