This review provides overall support for the dictum that legally referred clients do as well or better than voluntary clients in and after treatment; however, the review also shows some divergence in findings that have significance. The authors propose that the majority of the variation in coerced treatment outcomes is due to inconsistent terminologies for referral status, neglected emphasis on internal motivation, and infidelity in program implementation. Anglin and Hser (1991) recommend four considerations for the design and implementation of programs that serve legally coerced clients. First, the period of intervention should be lengthy, since drug dependence is a chronic, recurring condition. Prior research suggests an ideal treatment of 3 to 9 months, and several episodes of primary treatment, aftercare, and relapse should be expected. Second, treatment programs should provide a high level of structure, particularly during the early stages. Third, programs must be flexible; among community-based treatment clients, occasional drug use that does not appear to disrupt the overall recovery process should be handled on a client-by-client basis. Fourth, programs must undergo regular evaluation to determine their level of effectiveness and to detect changes in the client population they serve. 41 references