Prevention of type 2 diabetes by intensive lifestyle intervention designed to achieve and maintain ideal body weight was assessed in subjects with impaired glucose tolerance (IGT). Male subjects with IGT recruited from health-screening examinees were randomly assigned in a 4:1 ratio to a standard intervention group (control group) and intensive intervention group (intervention group). The final numbers of subjects were 356 and 102, respectively. The subjects in the control group and in the intervention group were advised to maintain body mass index (BMI) of <24.0 kg/m2 and of <22.0 kg/m2, respectively, by diet and exercise. In the intervention group, detailed instructions on lifestyle were repeated every 3-4 months during hospital visits. Diabetes was judged to have developed when two or more consecutive fasting plasma glucose (FPG) values exceeded 140 mg/dl. A 100g oral glucose tolerance test was performed every 6 months to detect improvement of glucose tolerance. The subjects were seen in an ordinary outpatient clinic. The cumulative 4-year incidence of diabetes was 9.3% in the control group, versus 3.0% in the intervention group, and the reduction in risk of diabetes was 67.4% (P < 0.001). Body weight decreased by 0.39 kg in the control group and by 2.18 kg in the intervention group (P < 0.001). The control group was subclassified according to increase and decrease in body weight. The incidence of diabetes was positively correlated with the changes in body weight, and the improvement in glucose tolerance was negatively correlated. Subjects with higher FPG at baseline developed diabetes at a higher rate than those with lower FPG. Higher 2h plasma glucose values and higher BMI values at baseline were also associated with a higher incidence of diabetes, but the differences were not significant. Subjects with a low insulinogenic index (DeltaIRI/DeltaPG 30 min after an oral glucose load) developed diabetes at a significantly higher rate than those with a normal insulinogenic index. Comparison of the BMI data and incidence of diabetes in five diabetes prevention studies by lifestyle intervention revealed a linear correlation between the incidence of diabetes and the BMI values, with the exception of the DaQing Study. However, the slope of the reduction in incidence of diabetes in the intensive intervention groups was steeper than expected simply on the basis of the reduction of BMI, suggesting that the effect of lifestyle intervention cannot be solely ascribed to the body weight reduction. We conclude that lifestyle intervention aimed at achieving ideal body weight in men with IGT is effective and can be conducted in an outpatient clinic setting.