A nationally representative sample of 326 nonmethadone outpatient drug abuse treatment organizations was surveyed in 1990, with the unit of analysis being the treatment organization. The outcome measure was the percentage of clients who completed or dropped out of treatment. The study looked at the percentage of black clients treated by the organization, the socioeconomic status of the organization's area, organizational factors, and treatment practices. Multivariate regression analysis was used to assess organizational rates of treatment success and failure among clients. Ordinary least squares regression analysis showed that race was not a predictor of treatment success once environmental factors were included in the analysis. The authors conclude that treatment organizations should be wary of defining persons for treatment in terms of their ethnic group membership and that treatment approaches should place more emphasis on socioenvironmental influences to which clients are exposed. 60 references, 3 notes, and 3 tables