• The efficacy of three different residential therapeutic communities for male veterans addicted to heroin was studied, comparing 181 subjects who were randomly assigned to one of the communities with 166 subjects briefly hospitalized only for the treatment of withdrawal symptoms. At two-year follow-up, subjects from both a professionally staffed community and a peer confrontation community were found significantly more likely than the withdrawal-only group to be working or attending school and less likely to have been convicted of a crime. An eclectic program employing both professionals and paraprofessionals was not found to exceed the withdrawal-only group on any of the major outcome variables. The two relatively successful communities, although different in structure and style, were both perceived by their residents to have greater program clarity, order, staff control, and orientation to personal problems than the unsuccessful program.