The relationship between skeletal muscle perfusion, interstitial glucose concentration and sarcolemmal permeability to glucose in exercise-induced increases in muscle insulin sensitivity is not well established. In 10 young men, a euglycemic hyperinsulinemic clamp was initiated 4 h after 1-legged exercise, and leg blood flow (LBF), leg glucose uptake (LGU) and muscle interstitial glucose concentration were measured. During the clamp LGU and LBF were higher (P<0.05) in the previously exercised than the control leg whereas the interstitial glucose concentration decreased to lower (P<0.05) values in the exercised (~3.1mM) than the control (~4.8mM) leg. Apparent sarcolemmal glucose permeability was twice as high (P<0.05) in the exercised compared with the rested leg. The NOS inhibitor L-NMMA decreased LBF in both legs and interstitial glucose concentration dropped to ~2.3 mM in the exercised but only to ~3.7 mM in non-exercised muscle. This abrogated the augmented effect of insulin on LGU in the exercised leg while apparent sarcolemmal permeability to glucose remained unchanged with L-NMMA in both legs. Doubling LBF by local infusion of ATP increased LGU in both legs without any major change in interstitial glucose concentration or sarcolemmal permeability to glucose. These findings suggest that during flow restriction due to L-NMMA, the interstitial glucose concentration becomes limiting for insulin-stimulated glucose uptake in previously exercised but not in non-exercised muscle. Therefore, the vasodilatory effect of insulin is an important component of the increased insulin sensitivity to stimulate glucose uptake following exercise by limiting the drop in the interstitial glucose concentration that occurs due to the increased sarcolemmal permeability to glucose.