Insulin binding to monocytes and insulin action in vivo was examined in 14 obese subjects during the postabsorptive state and after starvation and refeeding. Tissue sensitivity to insulin was evaluated with the euglycemic insulin clamp technique. The plasma insulin concentration is acutely raised and maintained 100 muU/ml above the fasting level, and plasma glucose is held constant by a variable glucose infusion. The amount of glucose infused is a measure of tissue sensitivity to insulin and averaged 285+/-15 mg/m(2) per min in controls compared to 136+/-13 mg/m(2) per min in obese subjects (P <0.001). (125)I-Insulin binding to monocytes averaged 8.3+/-0.4% in controls vs. 4.6+/-0.5% in obese subjects (P < 0.001). Insulin binding and insulin action were highly correlated in both control (r = 0.86, P < 0.001) and obese (r = 0.94, P < 0.001) groups. Studies employing tritiated glucose to measure glucose production indicated hepatic as well as extrahepatic resistance to insulin in obesity. After 3 and 14 days of starvation, insulin sensitivity in obese subjects decreased to 69+/-4 and 71+/-7 mg/m(2) per min, respectively, whereas (125)I-insulin binding increased to 8.8+/-0.7 and 9.0+/-0.4%. In contrast to the basal state, there was no correlation between insulin binding and insulin action. After refeeding, tissue sensitivity increased to 168+/-14 mg/m(2) per min (P < 0.001) whereas insulin binding fell to 5.0+/-0.3%. We conclude that (a) in the postabsorptive state insulin binding to monocytes provides an index of in vivo insulin action in nonobese and obese subjects and, (b) during starvation and refeeding, insulin binding and insulin action changes in opposite directions suggesting that postreceptor events determine in vivo insulin sensitivity.