One bout of exercise enhances insulin-stimulated glucose uptake (insulin action), but the effect is blunted by consumption of carbohydrate-containing food after exercise. The independent roles of energy and carbohydrate in mediating post-exercise insulin action have not been systematically evaluated in humans. The purpose of this study was to determine if varying carbohydrate availability, with energy intake held constant, mediates post-exercise insulin action. Ten young (21 +/- 2 y, overweight (body fat 37% +/- 3%) men and women completed 3 conditions in random order: (i) no-exercise (BASE), (ii) exercise with energy balance but carbohydrate deficit (C-DEF), and (iii) exercise with energy and carbohydrate balance (C-BAL). In the exercise conditions, subjects expended 30% of total daily energy expenditure on a cycle ergometer at 70% VO2 peak. Following exercise, subjects consumed a meal that replaced expended energy (~3000 kJ) and was either balanced (intake = expenditure) or deficient (-100 g) in carbohydrate. Twelve hours later, insulin action was measured by continuous infusion of glucose with stable isotope tracer (CIG-SIT). Changes in insulin action were evaluated using a one-way ANOVA with repeated measures. During CIG-SIT, non-oxidative glucose disposal (i.e., glucose storage) was higher in C-DEF than in BASE (27.2 +/- 3.2 vs. 16.9 +/- 3.5 micromol.L-1.kg-1.min-1, p < 0.05). Conversely, glucose oxidation was lower in C-DEF (8.6 +/- 1.3 micromol.L-1.kg-1.min-1) compared with C-BAL (12.2 +/- 1.2 micromol.L-1.kg-1.min-1), and BASE (17.1 +/- 2.2 micromol.L-1.kg-1.min-1), p < 0.05). Fasting fat oxidation was higher in C-DEF than in BASE (109.8 +/- 10.5 vs. 80.7 +/- 9.6 mg.min-1, p < 0.05). In C-DEF, enhanced insulin action was correlated with the magnitude of the carbohydrate deficit (r = 0.82, p < 0.01). Following exercise, re-feeding expended energy, but not carbohydrate, increased fasting fat oxidation, and shifted insulin-mediated glucose disposal toward increased storage and away from oxidation.