The BTC hypothesis is that a system of integrated testing, treatment, graduated responses, and supervision will reduce drug use in a defendant population. This, in turn, will reduce recidivism, improve social functioning and health status, and promote more effective use of system and treatment resources. The BTC model is being implemented in the adult justice systems of Birmingham, AL.; Jacksonville, FL.; and Tacoma, WA., as well as the juvenile justice system of Lane County (Eugene), OR. This report presents a summary of the lessons learned from BTC implementation to date, as identified at the Third BTC Site Meeting in Jacksonville on September 11-12, 2000. One of the lessons learned is that pre-existing issues affected BTC operation, such as jail overcrowding in Birmingham and Tacoma, heavy probation caseloads in each adult site, and minimally integrated data systems in each site. Another lesson is that the uncertainty about responsibilities was the main obstacle to collaboration. Other lessons are that judicial oversight under regular case processing was difficult to maintain; using and applying effective sanctions and incentives has been difficult; pretrial case management and drug monitoring components have been highly effective elements; and effective temporary alternatives to treatment placements must be established. Additional lessons identified are that effective data management and information exchange among BTC partners, the evaluators, and the National Institute of Justice (Federal BTC sponsor) are essential; and BTC's collaborative management model helped sites address other justice system issues.