Fluoxetine, an inhibitor of serotonin re-uptake, has been shown to cause weight loss in humans and animals. In order to determine the effects in diabetic subjects, 48 male and female, obese, type 2 non-insulin dependent diabetics being treated with insulin were randomized to receive fluoxetine 60 mg or placebo once daily in double blind fashion for 24 weeks. In all subjects, this treatment was preceded by four weeks and followed by six weeks of single blind placebo washout treatment. Subjects performed daily home glucose monitoring and were given instruction in a 1200 kcal American Diabetes Association diet. Fluoxetine treated subjects who completed the trial (n = 16) lost more weight than placebo treated subjects (n = 20) (9.3 +/- 2.4 vs. 1.9 +/- 2.9 kg +/- s.e.m, P less than 0.05). Subjects in the fluoxetine group also showed a greater percentage decrease in insulin dose than those in the placebo group (46.9 +/- 7.6% vs. 19.3 +/- 7.6%, P less than 0.01). During active treatment, the change in serum glucose levels did not differ between the two groups, while glycohemoglobin fell more in fluoxetine treated subjects than in placebo treated subjects at two of four follow-up visits. These results suggest that fluoxetine may be of benefit in the treatment of obese patients with type 2 non-insulin dependent diabetes mellitus.