A meta-analysis was conducted on studies using a treatment-comparison group design to evaluate HIV/AIDS risk-reduction interventions for clients enrolled in drug abuse treatment programs. Overall, the interventions studied were found to have a reliable positive (weighted) effect size (d = 0.31), and this was unlikely to be due to publication bias. Effect sizes for specific categories of outcome variables were 0.31 for knowledge, attitudes, and beliefs; 0.26 for sexual behavior; 0.62 for risk-reduction skills; and 0.04 for injection practices. A number of potential moderators were examined. Effect sizes were negatively correlated with the presence of predominantly ethnic minority samples and positively correlated with the number of intervention techniques used, the intensity of the intervention, intervention delivery at a later stage of drug treatment or within methadone treatment, and the presence of a number of specific intervention techniques.