Whether the practice of referral outside the school setting or direct intervention in the classroom as advocated by Patterson results in more efficient treatment effects is an empirical question. The study reported in this paper represents an initial effort to resolve this issue. The Social Learning Project has conducted treatemtn of sociall aggressive elementary school boys who are out of control in both school and home settings. When a referral is received from the school, intervention is typically begun in the home and onl initiated in the school when the parents have achieved behavior control at home. A study was designed to determine the relationship between treatment of socially deviant behavior in the home and in the classroom. Concurrent observations of deviant behavior were obtained in both settings and the results were compared for a small sample of subjects where it was possible to delay classroom intervention. Evidence predicts that dramatic treatment effects in one setting result in an increase in social deviancy in the non-treated setting as soon as the discrimination in schedules of reinforcement between setttings is learned. Evaluation of the generalization of treatment effects across all 13 deviant behaviors observed showed no changes in the rates of all deviant behaviors for the subjects in this study as a function of the 6-week home intervention.