We determined the effect of insulin on muscle blood flow and glucose uptake in humans using [15O]H2O, [18F]fluoro-2-deoxy-D-glucose ([18F]FDG), and positron emission tomography (PET). Femoral muscle blood flow was measured in 14 healthy volunteers (age 34 +/- 8 years, BMI 24.6 +/- 3.4 kg/m2 [means +/- SD]) before and at 75 min during a 140-min high-dose insulin infusion (serum insulin 2,820 +/- 540 pmol/l) under normoglycemic conditions. A dynamic scan of the femoral region was performed using PET for 6 min after injection of [15O]H2O to determine the 15O concentration in tissue. Regional femoral muscle blood flow was calculated using an autoradiographic method from the dynamic data obtained with PET and [15O]H2O. Femoral muscle glucose uptake was measured during hyperinsulinemia immediately after the flow measurement using PET-derived [18F]FDG kinetics and a three-compartment model. Whole-body glucose uptake was quantitated using the euglycemic insulin clamp technique. In the basal state, 84 +/- 8% of blood flow was confined to skeletal muscle. Insulin increased leg blood flow from 29 +/- 14 to 54 +/- 29 ml x kg-1 leg x min-1 (P < 0.001) and muscle flow from 31 +/- 18 to 58 +/- 35 ml x kg-1 muscle x min-1 (P < 0.005). Under insulin-stimulated conditions, 81 +/- 8% of blood flow was in muscle tissue (NS versus basal). Skeletal muscle explained 70 +/- 25% of the increase in leg blood flow. No correlation was observed between blood flow and glucose uptake when analyzed individually in identical regions of interest within femoral muscles. These data demonstrate that skeletal muscle accounts for most of the insulin-induced increase in blood flow. Insulin-stimulated rates of blood flow and glucose uptake do not colocalize in the same regions of muscle tissue, suggesting that insulin's hemodynamic and metabolic effects are differentially regulated.