Background: Few studies have compared long-term moderate-intensity aerobic versus light-resistance training on serial improvements in glucose tolerance in older people. Methods: Healthy, inactive older (74 ± 5 [SD] years) women (N = 20) were randomized into either a high-volume, moderate-intensity aerobic (ATM, n = 12) or a lower-intensity resistance training (RTL, n = 8) group. Both groups exercised under supervision 4 times per week for 45- to 60-min sessions over 9 months. Measurements of plasma glucose, insulin, and free fatty acid (FFA) responses to an oral glucose tolerance test (OGTT) were performed at baseline and at 3, 6, and 9 months 48 hours after the last exercise session. Results: We observed significant improvements in 2-hour glucose concentrations at 3, 6, and 9 months among women in the RTL(152 ± 42 vs 134 ± 33 vs 134 ± 24 vs 130 ± 27 mg · dL−1; P < .05), but not the ATM(151 ± 25 vs 156 ± 37 vs 152 ± 40 vs 155 ± 39 mg · dL−1) group. These improvements were accompanied by an 18% (P < .07) decrease in basal FFA concentrations in the RTLgroup, whereas basal and 30-min FFA concentrations increased (P < .05) after training in the ATMgroup. Conclusions: These findings suggest that the net physiological benefits of exercise might have been blunted in the ATMgroup, owing to higher circulating levels of FFA, which might have temporarily interfered with insulin action.