We investigated the efficacy of additional administration of 400 mg troglitazone (+T), which became available as a treatment for type 2 diabetes following the demonstration of its ability to reduce insulin resistance, in combination with diet (D + T) or sulfonylurea (S + T) therapy. Body fat area as determined by computed tomographic (CT) scanning at the umbilical level, as well as several clinical and biochemical parameters of glycemic control and lipid metabolism, were compared before and after 3 months of additional treatment with troglitazone. The body mass index (BMI) tended to increase in both groups (22.7 +/- 0.6 v 23.2 +/- 0.6 kg/m2 in D + T, nonsignificant [NS]; 22.2 +/- 0.5 v 22.3 +/- 0.5 kg/m2 in S + T, NS), while it tended to decrease in the control group (only diet therapy, 23.6 +/- 0.6 v 23.1 +/- 0.8 kg/m2, NS). Mean blood pressure ([BP] 96 +/- 3 v 89 +/- 4 mm Hg, P < .05) decreased significantly in the D + T group. Changes in the glycemic and lipid profile and leptin did not reach statistical significance. The D + T group showed a significant decline in immunoreactive insulin ([IRI] 12.4 +/- 1.2 v 8.0 +/- 1.0 microU/mL, P < .05), reflecting markedly reduced insulin resistance, as well as a significant increase in plasma insulin-like growth factor-1 ([IGF-1] 175.7 +/- 14.2 v 189.8 +/- 12.6 ng/mL, P < .05). A slight weight gain was associated with a tendency for subcutaneous fat to increase, while visceral fat decreased in both troglitazone-treated groups. The decrease in the visceral to subcutaneous fat ratio (V/S ratio) was statistically significant in the D + T group (1.09 +/- 0.11 v 0.94 +/- 0.09, P < .05), while the V/S ratio in the control group did not change. A notable finding of this study is the difference in the response to troglitazone between subcutaneous and visceral adipose tissue. It is suggested that troglitazone may exert beneficial effects by reducing visceral fat.