This study was designed to investigate the effect of exercise intensity on insulin resistance by comparing moderate- and high-intensity interventions of equal energy cost. Maximum oxygen consumption (VO_2max), insulin, glucose and triglycerides were measured in 64 sedentary men before random allocation to a non-exercise control group, a moderate-intensity exercise group (three 400-kcal sessions per week at 60% of VO_2max) or a high-intensity exercise group (three 400-kcal sessions per week at 80% of VO_2max). An insulin sensitivity score was derived from fasting concentrations of insulin and triglycerides, and insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR). Data were available for 36 men who finished the study. After 24 weeks, insulin concentration decreased by 2.54±4.09 and 2.37±3.35 mU l⁻¹, insulin sensitivity score increased by 0.91±1.52 and 0.79±1.37, and HOMA-IR decreased by −0.6±0.8 and −0.5±0.8 in the moderate- and high-intensity exercise groups, respectively. When data from the exercise groups were combined, one-way analysis of variance with one-tailed post hoc comparisons indicated that these changes were significantly greater than those observed in the control group (all P<0.05). Twenty-four week changes in insulin concentration, insulin sensitivity score and HOMA-IR were not significantly different between the exercise groups. These data suggest that exercise training is accompanied by a significant reduction in insulin resistance, as indicated by well-validated surrogate measures. These data also suggest that moderate-intensity exercise is as effective as high-intensity exercise when 400 kcal are expended per session.