Background: Patients with depression often report impairments in social functioning. From a patient perspective, improvements in social functioning might be an important outcome in psychotherapy for depression. Therefore, it is important to examine the effects of psychotherapy on social functioning in patients with depression. Method: We conducted a meta-analysis on studies of psychotherapy for depression that reported results for social functioning at post-treatment. Only studies that compared psychotherapy to a control condition were included (31 studies with 2956 patients). Results: The effect size of psychotherapy on social functioning was small to moderate, before [Hedges' g = 0.46, 95% confidence interval (CI) 0.32-0.60] and after adjusting for publication bias (g = 0.40, 95% CI 0.25-0.55). Univariate moderator analyses revealed that studies using care as usual as a control group versus other control groups yielded lower effect sizes, whereas studies conducted in the USA versus other countries and studies that used clinician-rated instruments versus self-report yielded higher effect sizes. Higher quality studies yielded lower effect sizes whereas the number of treatment sessions and the effect size of depressive symptoms were positively related to the effect size of social functioning. When controlling for these and additional characteristics simultaneously in multivariate meta-regression, the effect size of depressive symptoms, treatment format and number of sessions were significant predictors. The effect size of social functioning remained marginally significant, indicating that improvements in social functioning are not fully explained by improvements in depressive symptoms. Conclusions: Psychotherapy for depression results in small to moderate improvements in social functioning. These improvements are strongly associated with, but not fully explained by, improvements in depressive symptoms. Copyright © Cambridge University Press 2014.