W e tested the hypothesis that endothelium-dependent vasodilatation is a determinant of insulin resistance of skeletal muscle glucose uptake in human obesity. Eight obese (age 26+/-1 yr, body mass index 37+/-1 kg/m2) and seven nonobese males (25+/-2 yr, 23+/-1 kg/m2) received an infusion of bradykinin into the femoral artery of one leg under intravenously maintained normoglycemic hyperinsulinemic conditions. Blood flow was measured simultaneously in the bradykinin and insulin- and the insulin-infused leg before and during hyperinsulinemia using [15O]-labeled water ([15O]H2O) and positron emission tomography (PET). Glucose uptake was quantitated immediately thereafter in both legs using [18F]- fluoro-deoxy-glucose ([18F]FDG) and PET. Whole body insulin-stimulated glucose uptake was lower in the obese (507+/-47 mumol/m2 . min) than the nonobese (1205+/-97 micromol/m2 . min, P < 0.001) subjects. Muscle glucose uptake in the insulin-infused leg was 66% lower in the obese (19+/-4 micromol/kg muscle . min) than in the nonobese (56+/-9 micromol/kg muscle . min, P < 0.005) subjects. Bradykinin increased blood flow during hyperinsulinemia in the obese subjects by 75% from 16+/-1 to 28+/-4 ml/kg muscle . min (P < 0.05), and in the normal subjects by 65% from 23+/-3 to 38+/-9 ml/kg muscle . min (P < 0.05). However, this flow increase required twice as much bradykinin in the obese (51+/-3 microg over 100 min) than in the normal (25+/-1 mug, P < 0.001) subjects. In the obese subjects, blood flow in the bradykinin and insulin-infused leg (28+/-4 ml/kg muscle . min) was comparable to that in the insulin-infused leg in the normal subjects during hyperinsulinemia (24+/-5 ml/kg muscle . min). Despite this, insulin-stimulated glucose uptake remained unchanged in the bradykinin and insulin-infused leg (18+/-4 mumol/kg . min) compared with the insulin-infused leg (19+/-4 micromol/kg muscle . min) in the obese subjects. Insulin-stimulated glucose uptake also was unaffected by bradykinin in the normal subjects (58+/-10 vs. 56+/-9 micromol/kg . min, bradykinin and insulin versus insulin leg). These data demonstrate that obesity is characterized by two distinct defects in skeletal muscle: insulin resistance of cellular glucose extraction and impaired endothelium-dependent vasodilatation. Since a 75% increase in blood flow does not alter glucose uptake, insulin resistance in obesity cannot be overcome by normalizing muscle blood flow.