Bulimia nervosa represents a serious public health problem in the United States. We performed an 8-week, double-blind trial comparing fluoxetine hydrochloride (60 and 20 mg/d) with placebo in 387 bulimic women treated on an outpatient basis. Fluoxetine at 60 mg/d proved superior to placebo in decreasing the frequency of weekly binge-eating and vomiting episodes at end point. Fluoxetine at 20 mg/d produced an effect between that of the 60-mg/d dosage and that of placebo. Depression, carbohydrate craving, and pathologic eating attitudes and behaviors also improved significantly with fluoxetine, with the higher dosage again showing a more robust effect than the lower dosage. Several adverse events (ie, insomnia, nausea, asthenia, and tremor) occurred significantly more frequently with fluoxetine (60 or 20 mg/d) than with placebo. However, there was no statistically significant difference among treatment groups in the proportion of patients discontinuing the study because of adverse events.