Using the glucose clamp technique, glucose uptake was determined isotopically in normal human volunteers at plasma glucose concentrations of congruent to 60, 95, and 160 mg/dl during insulin infusions that increased plasma insulin to congruent to 20, 80, and 160 microU/ml. Because glucose uptake was found to be a linear function of plasma insulin at each plasma glucose concentration (r greater than 0.92, P less than 0.01), glucose uptake at 0 plasma insulin was estimated by linear regression analysis. The values thus derived (1.30, 1.62, and 2.59 mg . kg-1 . min-1 for plasma glucose concentrations of 60, 95, and 160 mg/dl, respectively) produced a linear Eadie-Hofstee plot, suggesting that insulin-independent glucose uptake followed Michaelis-Menten kinetics. The Km for glucose uptake at 0 plasma insulin (congruent to 10 mM) was similar to those observed for glucose uptake at the other plasma insulin concentrations studied (congruent to 9-12 mM), but its Vmax was less (5.2 vs. 6.4, 18.5, and 26.8 mg . kg-1 . min-1 for congruent to 20, 80, and 160 U/ml, respectively). These results indicate that in postabsorptive human subjects 75-85% of glucose uptake is noninsulin-mediated and provide additional support for the concept that insulin may increase glucose uptake merely by providing additional transport sites. The method described herein provides an assessment of insulin-independent glucose uptake in vivo that may prove useful in distinguishing between intrinsic defects of the glucose transport system and those due to defects in insulin action.