This study analyzed the Alcoholics Anonymous participation and the impact on the abstinence rate of 103 alcohol dependent patients (ICD-10) during the 24 months after their discharge from high-structured out-patient treatment. The treatment retention amounted to n = 74 (72 %), 18 of the 25 dropouts took place because of alcohol relapse. At 6-, 12- and 24-months-follow-up 87 - 95 % of the patients were successfully located and re-interviewed. Analyses revealed that 64 % of the patients were abstinent at the 6-months-follow-up evaluation, 56 % at the total 12-months-follow-up evaluation. 49 % of the patients remained abstinent until the 24-months-follow-up evaluation, 14 % were improved and 37 % relapsed. 56 patients (55 %) participated in selfhelp-groups the first six months following treatment, two years later 45 patients (44 %) still attended a group. 53 - 56 % participated on a weekly basis. Patients who participated regularly on a weekly basis in self-help-groups fared the best on 2-year outcome. Patients who infrequently at all participated or not had the poorest outcome (relapse). After controlling for confounding variables (sex, treatment drop-out, relapse during treatment) these results were still statistical significant. Results indicate the predictive value of AA attendance for relapse prevention after controlling for confounding variables. The value of self-help-groups in the network of alcoholism treatment is discussed.