Insulin action and obesity are both correlated with the density of muscle capillary supply in humans. Since the altered muscle anatomy in the obese might affect interstitial insulin concentrations and reduce insulin action, we have cannulated peripheral lymphatic vessels in lean and obese males, and compared peripheral lymph insulin concentrations with whole body glucose uptake during a euglycemic, hyperinsulinemic clamp. Lymph insulin concentrations in the lower limb averaged only 34% of arterial insulin concentrations during 150 min of insulin infusion. Obese subjects had the highest arterial (P < or = 0.0001) and lymph insulin (P < 0.005) concentrations, but the lowest glucose uptake rates (P < 0.002). In contrast to the initial steep rise then plateau of arterial insulins, both lymph insulin and whole body glucose uptake rates rose slowly and did not consistently reach a plateau. In each individual, the glucose uptake closely correlated with peripheral lymphatic insulin concentrations (mean r2 = 0.95). The coupling between glucose uptake and lymph insulin (glucose uptake/pmol insulin) was much steeper in lean subjects than in the obese (P < or = 0.0001). These results indicate that even if insulin diffusion into tissues is rate limiting for insulin action, a tissue defect rather than an insulin diffusion defect causes insulin resistance in obese subjects.