The present investigation examined the significance of a past history of substance abuse on treatment outcome for bulimia nervosa. Seventy-five women with bulimia nervosa participated in a 6-week double-blind outpatient trial of desipramine; 19 patients had a history of substance abuse and 56 did not. Although patients with a history of substance abuse reported higher levels of anxiety and depression at presentation for treatment than patients without such histories, the two groups reported a similar age of onset of their bulimia nervosa and similar severity of eating pathology with regards to binge and vomit frequencies and measures of concern about body shape and weight. On all outcome measures, the improvement of the substance abuse group was equal to or greater than that in the group without a history of substance abuse. Results suggest that a past history of substance abuse has no implication regarding response to pharmacological treatment for bulimia nervosa.