The authors present findings from the first investigation of the use of alcohol, nicotine, and caffeine in nonsubstance-abusing outpatients with major depressive disorder. The patients (N = 94) were assessed for their intake of alcohol, nicotine, and caffeine, and then treated openly for 8 weeks with 20 mg/day of fluoxetine. The degree of alcohol consumption at baseline was a significant predictor of poorer response to the antidepressant. This relationship remained significant even after adjusting for severity of depression at baseline. Even moderate levels of alcohol consumption appear to negatively affect pharmacologic treatment in depressed outpatients.