(i) To evaluate the effect of receiving one of two brief interventions in reducing alcohol consumption among general hospital patients compared with usual care. (ii) To assess whether a brief intervention of self-efficacy enhancement was superior to a self-help booklet in reducing alcohol consumption. A three-arm cluster randomized controlled trial. Seven general medical, six general surgical, one dermatology and two otolaryngology wards of a large teaching hospital covering a large urban and rural area. A total of 215 of 789 in-patients aged 18-75 years, who screened positive for alcohol consumption in excess of national recommended limits according to a 7-day retrospective drinking diary. Participants were allocated to receive one of three interventions: (i) face-to-face self-efficacy enhancement; (ii) a self-help booklet; or (iii) usual care. The primary outcome measure was change in reported alcohol consumption at 6-month follow-up as measured by a 7-day retrospective drinking diary. Secondary outcomes were change in: number of alcohol drinking days in last week; the maximum units of alcohol consumed on any one day in last week; and Drinking Refusal Self-efficacy Expectancy Questionnaire score. Compared to the usual care group the self-efficacy enhancement group (-10.1 units 95% CI -16.1 to -4.1) and the self-help booklet group (-10.0 units 95% CI -16.0 to -3.9) had greater reductions in self-reported weekly alcohol consumption. There was no evidence that self-efficacy enhancement was superior to the self-help booklet (P = 0.96). Brief interventions delivered in hospital offer simple means of helping heavy drinkers to reduce their alcohol consumption.