The existence of an overall association between severity of depression and level of social functioning is well documented. To increase the probability of a long-term recovery, a normal level of social functioning is essential. It is currently unknown whether combined therapy has a better outcome than pharmacotherapy with regard to social functioning. In a 6-month randomized clinical trial in outpatients with major depression, all patients studied had a baseline score of at least 14 points on the 17-item Hamilton Depression Rating Scale (HDRS). The two conditions consist of pharmacotherapy (PhT) (N=84) and combined therapy (CoT), pharmacotherapy plus 16 sessions of short psychodynamic supportive psychotherapy (N=83). Efficacy was assessed using the 17-item HDRS, the Clinical Global Impression (CGI) Severity and Improvement scales, the Depression subscale of the Symptom Checklist-90 (SCL-90), the Quality of Life Depression Scale (QLDS), more the Groningen Social Disability Schedule (GSDS). Severity of depression decreased significantly (on the SCL-90 Depression subscale and the QLDS) more in the CoT condition. A larger improvement in social functioning was demonstrated for remitted patients than for nonremitted patients. The number of dimensions of social functioning that had improved significantly was higher in CoT than in PhT. There was a moderate advantage of the CoT condition on both depressive symptoms and level of social functioning in comparison with PhT. We also found a positive association between depression severity and level of social functioning.