To assess the mechanisms of the insulin resistance in human obesity, we have determined, using a modification of the euglycemic glucose clamp technique, the shape of the in vivo insulin-glucose disposal dose-response curves in 7 control and 13 obese human subjects. Each subject had at least three euglycemic studies performed at insulin infusion rates of 15, 40, 120, 240, or 1,200 mU/M2/min. The glucose disposal rate was decreased in all obese subjects compared with controls (101 +/- 16 vs. 186 +/- 16 mg/M2/min) during the 40 mU/M2/min insulin infusion. The mean dose-response curve for the obese subjects was displaced to the right, i.e., the half-maximally effective insulin concentration was 270 +/- 27 microU/ml for the obese compared with 130 +/- 10 microU/ml for controls. In nine of the obese subjects, the dose-response curves were shifted to the right, and maximal glucose disposal rates (at a maximally effective insulin concentration) were markedly decreased, indicating both a receptor and a postreceptor defect. On the other hand, four obese patients had right-shifted dose-response curves but reached normal maximal glucose disposal rates, consistent with decreased insulin receptors as the only abnormality. When the individual data were analyzed, it was found that the lease hyperinsulinemic, least insulin-resistant patients displayed only the receptor defect, whereas those with the greatest hyperinsulinemia exhibited the largest post-receptor defect, suggesting a continuous spectrum of defects as one advances from mild to severe insulin resistance. When insulin's ability to suppress hepatic glucose output was assessed, hyperinsulinemia produced total suppresssion in all subjects. The dose-response curve for the obese subjects was shifted to the right, indicating a defect in insulin receptors. Insulin binding to isolated adipocytes obtained from the obese subjects was decreased, and a highly significant inverse linear relationship was demonstrated between insulin binding and the serum insulin concentration required for halfmaximal stimulation of glucose disposal. In conclusion: (a) decreased cellular insulin receptors contribute to the insulin resistance associated with human obesity in all subjects; (b) in the least hyperinsulinemic, insulin-resistant patients, decreased insulin receptors are the sole defect, whereas in the more hyperinsulinemic, insulin-resistant patients, the insulin resistance is the result of a combination of receptor and postreceptor abnormalities; (c) all obese patients were insensitive to insulin's suppressive effects on hepatic glucose output; this was entirely the result of decreased insulin receptors; no postreceptor defect in this insulin effect was demonstrated.