To test the hypothesis that depression is significantly underdiagnosed in general medical settings, the Zung Self-Rating Depression Scale was administered to 1,086 family medicine outpatients seen during a 12-month period before their initial medical examination. The effects of such screening on clinical recognition and treatment of depression were examined. Of the 1,086 patients, 143 (13.2%) were symptomatically depressed. These patients were randomized into two groups: 102 were identified as clinically depressed to their physician, and the remaining 41 were not (control group). Physicians diagnosed depression in 15% of the control group and in 68% of the identified group. At 4-week follow-up, 64% of the identified patients who were treated with maprotiline (16 of 25) showed improvement; only 28% of the identified patients who were electively not treated improved. Improvement occurred in only 18% of the control group. It appears that the diagnosis of depression is not ordinarily made in family medicine outpatient settings and that self-rating depression scales are useful diagnostic aids, whose regular use is indicated by the high prevalence of depression in general medical populations.