BACKGROUND: British Pakistani women have a high prevalence of depression. There are no reported psychosocial interventions for depression in ethnic minorities in the UK. AIMS: To determine the efficacy of a social group intervention compared with antidepressants, and whether the combination of the two is more efficacious than either alone. METHOD: A total of 123 women with depression participated in the primary care-based cluster randomised controlled trial (ISRCTN19172148). Outcome measures were severity of depression (Hamilton Rating Scale for Depression), social functioning and satisfaction at 3 and 9 months. RESULTS: Greater improvement in depression in the social intervention group and the combined treatment group compared with those receiving antidepressants alone fell short of significance. There was significantly greater improvement in social functioning in the social intervention and combined treatment groups than in the antidepressant group at both 3 and 9 months. CONCLUSIONS: Pakistani women with depression found the social groups acceptable and their social function and satisfaction improved if they received social treatment compared with the receipt of antidepressants alone.