This document, presented at a symposium on violence and aggression in America, reports a longitudinal study of environmental influences in the development of aggressive behavior. Subjects were 875 third graders when originally assessed with a peer nomination technique in 1960. High, moderate, and low degrees of aggression were identified in the population, and information was collected on the following kinds of variables: (1) instigators, (2) reinforcers, (3) identification, and (4) socio-cultural. Each of these factors was significantly related to aggression in the original data analysis. In 1970, 427 of the original subjects were located and reinterviewed. The sample was overloaded with low aggression youths, indicating a relationship between residential mobility and aggressive behavior. The second assessment included retesting with the peer rating, self-reports to determine extent of aggressive habits, and administration of the MMPI. Analysis of results indicates the stability of aggressive behavior. At age 19, subjects' aggression was predicted only by identification and socio-cultural variables. Significant sex differences were apparent. Discussion focuses on socialization factors in the development of aggression, particularly sex role development and the influence of modeling through television. (DP)