The current study examined the insulin and glucose response during an oral glucose tolerance test (OGTT) in overweight young adults prior to and following exercise training in the Midwest Exercise Trial (MET). Subjects (N = 66) were randomly assigned to non-exercise control (CON; 16 females, 13 males) or exercise (EX; 22 females, 15 males) groups. EX performed supervised and verified exercise on 3 to 5 days per week in 20- to 45-minute sessions at 60% to 75% of heart rate reserve. OGTTs and assessments for body mass, body composition, and maximal oxygen consumption (Vo2 max) were performed at baseline, and after 9 and 16 months of training. Blood was collected during a 75-g OGTT and analyzed for glucose and insulin concentrations with the total area under the glucose and insulin curves used in the analysis. The EX males had significant decreases from baseline to 9 months in body mass (94.8 ± 12.5 to 89.2 ± 9.8 kg) and percent fat (28.3 ± 4.8 to 24.2 ± 3.9) with no further changes at 16 months. CON females had significant increases in body mass (78.2 ± 6.4 to 81.1 ± 8.1 kg) and percent fat (36.6 ± 4.2 to 37.8 ± 4.7) from baseline to 16 months. Vo2 max increased significantly from baseline to 9 months in the EX males (3.67 ± 0.62 to 4.36 ± 0.55 L/min) and EX females (2.53 ± 0.32 to 2.99 ± 0.42 L/min). For glucose area under the curve, there were no significant differences between EX or CON across the 16 months of the study. For insulin area under the curve, there was a significant decrease for male EX from baseline to 9 months (12,535 ± 6,114 to 8,390 ± 4,231 μU/L/180 min). We conclude that regular exercise in healthy, previously sedentary overweight adult males leads to improvements in Vo2 max and weight loss and a reduction in the insulin concentration required to dispose of a set glucose load. In females, improvement in Vo2 max without weight loss does not lead to improvement in insulin sensitivity. © 2003 Elsevier Inc. All rights reserved.