Aims: To examine the effectiveness of a brief intervention (BI) and cognitive behaviour therapy (CBT) for alcohol abuse. Design: A randomized trial with clients randomized within counsellors. Setting: Community-based drug and alcohol counselling in Australia. Participants: Of all new clients attending counselling, 869 (82%) completed a computerized assessment at their first consultation. Four hundred and twenty-one (48%) were defined as eligible, of whom 295 (70%) consented and were allocated randomly to an intervention. Of these, 133 (45%) were followed-up at 6 months post-test. Interventions: BI comprised the elements identified by the acronym FRAMES: feedback, responsibility, advice, menu, empathy, self-efficacy. Face-to-face counselling time was not to exceed 90 minutes. CBT comprised six consecutive weekly sessions: introduction; cravings and urges; managing crises; saying ‘no’ and solving problems; emergencies and lapses; and maintenance. Total face-to-face counselling time was 270 minutes (six 45-minute sessions). Measurements: Treatment outcomes are measured in terms of counsellor compliance, client satisfaction, weekly and binge consumption, alcohol-related problems, the AUDIT questionnaire and cost-effectiveness. Findings: When analysed on an intention-to-treat basis and for those followed-up, treatment outcomes between BI and CBT were not statistically significantly different at pre- or post-test, whether considered as continuous or categorical variables. BI was statistically significantly more cost-effective than CBT and there was no difference between them in clients’ reported levels of satisfaction. Conclusion: For low-dependence alcohol abuse in community settings, BI may be the treatment of choice.