SummaryForty‐six clients of a clinical psychology service were assigned, in accordance with a three‐step design which involved a random element, to abstinence (ABST) or controlled drinking (CD), and were followed up to 12 months after initial assessment. No support was found for the dependence hypothesis: those with higher levels of dependence/severity assigned to CD and those with lower levels assigned to ABST were just as likely as other clients to do well, and there was no relationship between level of dependence/severity and the type of drinking outcome (ABST or CD). There was tentative support for the hypothesis that experience in the follow‐up year would be related to initial beliefs and preferences, although these were not independent of treatment type. The present report is not able to comment on the prediction of outcome in the longer term.