Aims and design. A randomized clinical trial was performed to evaluate the influence of two formats of cognitive-behavioral psychotherapy (individual vs. group) in the treatment of alcohol and/or drug dependent patients. Setting. Public outpatient drug dependence service. Participants. One hundred and fifty-five alcohol and/or drug-dependent patients. Intervention. The patients were randomly assigned to individual ( n = 77) or group ( n = 78) treatment formats. The treatment was developed into two phases: acquisition (eight sessions) and maintenance (nine sessions), distributed over an 8-month period. Measurements. Alcohol and drug use, severity of dependence, and alcohol- and drug-related problems were evaluated at pre-treatment and 15 months after admission to treatment. Findings. At follow-up evaluation both groups of patients presented similar levels of drug consumption, dependence and associated problems. Although group-treated patients reported slightly higher levels of alcohol consumption (both at baseline and follow-up) differences between the formats disappear if baseline levels are included as covariates. Compliance with treatment and a measure of drug severity were predictors of success for the drug dependents. The number of sessions attended and high GGT levels at admission were positively correlated with success for the alcohol dependents. Conclusions. The two modalities presented similar outcomes and, as the group format could present a better cost-benefit ratio, it may be used without decreasing compliance with treatment or treatment effectiveness.