Only a few drug courts have been the subject of high-quality, careful evaluation research; consequently, evidence of the impact of drug courts on key outcomes, such as recidivism and drug use, remains largely anecdotal. This lack of scientific evidence of the impact of drug courts is particularly acute for juvenile drug courts, which have developed more recently than adult drug courts. Much of what is known about the effectiveness of juvenile drug courts must be extrapolated from research on adult drug courts. Despite the rarity of sound drug-court evaluations, the strengths and weaknesses of existing evaluations can be analyzed and conclusions drawn. This paper reviews existing research, documents research areas in the drug court field that need further examination, and suggests ways to strengthen drug court evaluations. The review of drug-court evaluations focused on all relevant outcome evaluations completed as of 2003. A total of 15 quasi-experimental (strong) evaluations were identified, along with 4 experimental evaluations. Only seven juvenile drug court evaluations that included client outcomes could be located for this analysis. The review concludes that research on drug courts has produced only limited evidence of their impact. This is due to problems in designing and supporting drug court evaluations, lack of attention to theory-based evaluations, and difficulty in generalizing from evaluations due to the variation in drug court operations. This review emphasizes the importance of rigorous, theory-based evaluation research for future policy and practice. 32 references