This paper focuses on implementation problems in randomized trials of outpatient psychosocial treatments for drug abusers. The authors examined these problems in nine clinical trial studies and drew on published literature and their own research experiences. Common problems faced by principal investigators include the need for midstream treatment protocol and research design modifications based on the response of both clients and clinical staff, tension between research and clinical requirements, and the need to administer a large, complex organization over a substantial period of time. Solutions include conducting a pilot study, employing advanced research and analysis methods that can incorporate complex design variations, fostering a team spirit between diverse staffs, and employing special management structures.