A review of controlled evaluations of adolescent and other drug abuse treatment programs concludes that some treatment is better than no treatment, that few comparisons of treatment method have consistently demonstrated the superiority of one method over another, that posttreatment relapse rates are high, and that more controlled studies of adolescent treatment which allow evaluation of the elements of treatment are needed. In the absence of the clear superiority of specific treatment techniques, it is suggested that examination of factors related to relapse can provide an empirical base for identifying effective treatment approaches. Pretreatment, during-treatment, and posttreatment factors related to relapse are reviewed from existing studies of adolescent treatment. Program components are suggested which have been associated with or hold promise for reducing factors associated with higher relapse rates and increasing factors associated with lower relapse rates.