AIMS: To examine the feasibility, acceptability and preliminary cost-effectiveness of a lay counsellor delivered psychological treatment for men with alcohol dependence in primary care. DESIGN: Single-blind individually randomized trial comparing counselling for alcohol problems (CAP) plus enhanced usual care (EUC) versus EUC only. SETTING: Ten primary health centres in Goa, India. PARTICIPANTS: Men (n = 135) scoring ≥ 20 on the Alcohol Use Disorder Identification Test (AUDIT). Sixty-six participants were randomized to EUC and 69 to CAP + EUC. INTERVENTIONS: CAP, a lay counsellor-delivered psychological treatment for harmful drinking, with referral to de-addiction centre for medically assisted detoxification. EUC comprised consultation with physician, providing screening results and referral to a de-addiction centre. MEASUREMENTS: Baseline socio-demographic data, readiness to change and perceived usefulness of counselling. Acceptability and feasibility process indicators such as data on screening and therapy. Outcomes were measured at 3 and 12 months post-randomization and included remission, mean daily alcohol consumed, percentage of days abstinent (PDA), percentage of days of heavy drinking (PDHD), recovery, uptake of detoxification services, impacts of alcohol dependence, resource use and costs. FINDINGS: Participants in the CAP + EUC arm had more numerically but not statistically significantly favourable outcomes compared with those in the EUC arm for (a) remission at 3 months [adjusted odds ratio (aOR) = 1.95, 95% confidence interval (CI) = 0.74-5.15] and 12 months (aOR = 1.90, 95% CI = 0.72-5.00), (b) proportion of non-drinkers at 3 months (aOR = 1.26; 95% CI = 0.58-2.75) and 12 months (aOR = 1.25; 95% CI = 0.58-2.64) and (c) ethanol consumption among drinkers at 3 months (count ratio = 0.91; 95% CI = 0.58-1.45) and 12 months (count ratio = 1.06; 95% CI = 0.73-1.54). There was no statistically significant evidence of a difference in the occurrence of serious adverse events between the two arms. From a societal perspective, there was a 53% chance of CAP + EUC being cost-effective in achieving remission at 12 months at the willingness-to-pay threshold of $415. CONCLUSIONS: Lay counsellor-delivered psychological treatment for men with alcohol dependence (AD) in primary care may be effective in managing AD in low- and middle-income countries. A definitive trial of the intervention is warranted.