Thirteen consecutive referrals of bulimic patients who met DSM-III criteria for bulimia were treated in an open-label, flexible-dose study with trazodone. Three of the 13 dropped out before the fourth week of treatment, the minimum duration of treatment for evaluable subjects, and hence were not included in the analyses. For the 10 evaluable patients, the mean duration of treatment was 6.9 weeks and the mean maximum dose of trazodone was 410 mg (range, 250-600 mg). The number of binge eating and vomiting episodes was significantly decreased (p = 0.05 and 0.06, respectively). These episodes were reduced to zero in four patients and by 55-99% in two patients. Carbohydrate cravings and urges to binge eat were significantly diminished in intensity (p less than 0.02 and 0.008, respectively). The total score (p = not significant) and three subscale scores (p = 0.04, 0.09, and 0.10) of the Eating Disorders Inventory decreased. The mean Hamilton Depression Scale score fell from 10.4 to 3.3 (p = 0.002). Only mild side effects were noted: five subjects complained of morning drowsiness and two of headache. Mean weight was essentially unchanged: pretreatment, 58.5 kg; posttreatment, 57.3 kg. The lack of weight gain represents an advantage of trazodone over other currently prescribed antidepressants, particularly for this group of patients whose fear of becoming fat is a part of their basic pathology.