W hether insulin-mediated vasodilation is important in determining insulin's overall action to stimulate glucose uptake is unknown. To this end, we measured leg glucose uptake during euglycemic hyperinsulinemic clamps performed at two insulin doses (40 mU/m2 per min, n = 6 and 120 mU/m2 per min, n = 15) alone and during a superimposed intrafemoral artery infusion of GN-monomethyl-L-arginine (L-NMMA) designed to blunt insulin-mediated vasodilation. During the higher dose study, hyperinsulinemia resulted in about a twofold rise in basal leg blood flow from 0.24 +/- 0.02 to 0.45 +/- 0.05 liter/min, P < 0.0001. L-NMMA infusion resulted in a net 21% reduction in leg glucose uptake from 114 +/- 18 mg/min to 85 +/- 13 mg/min, P < 0.001. We also found a significant relationship between the rate of insulin-stimulated whole body glucose uptake and the magnitude of flow dependent glucose uptake (r = 0.57, P = 0.02). Data obtained during the lower dose insulin infusion resulted in similar findings. In conclusion, in healthy lean subjects, insulin-stimulated muscle blood flow contributes to both insulin responsiveness and insulin sensitivity. The most insulin-sensitive subjects appear to be the most reliant on muscle perfusion for insulin action. Insulin-mediated vasodilation is an important physiological determinant of insulin action.