T he aim of this study was to determine the role of increased plasma cortisol levels in the pathogenesis of hypoglycemia-associated autonomic failure. Experiments were carried out on 16 lean, healthy, overnight fasted male subjects. One group (n = 8) underwent two separate, 2-d randomized experiments separated by at least 2 mo. On day 1 insulin was infused at a rate of 1.5 mU/kg per min and 2 h clamped hypoglycemia (53 +/- 2 mg/dl) or euglycemia (93 +/- 3 mg/dl) was obtained during morning and afternoon. The next morning subjects underwent a 2-h hyperinsulinemic (1.5 mU/kg per min) hypoglycemic (53 +/- 2 mg/dl) clamp study. In the other group (n = 8), day 1 consisted of morning and afternoon 2-h clamped hyperinsulinemic euglycemia with cortisol infused to stimulate levels of plasma cortisol occurring during clamped hypoglycemia (53 mg/dl). The next morning (day 2) subjects underwent a 2-h hyperinsulinemic hypoglycemic clamp identical to the first group. Despite equivalent day 2 plasma glucose and insulin levels, steady state epinephrine, norepinephrine, pancreatic polypeptide, glucagon, ACTH and muscle sympathetic nerve activity (MSNA) values were significantly (R < 0.01) blunted after day 1 cortisol infusion compared to antecedent euglycemia. Compared to day 1 cortisol, antecedent hypoglycemia produced similar blunted day 2 responses of epinephrine, norepinephrine, pancreatic polypeptide and MSNA compared to day 1 cortisol. Antecedent hypoglycemia, however, produced a more pronounced blunting of plasma glucagon, ACTH, and hepatic glucose production compared to day 1 cortisol. We conclude that in healthy overnight fasted men (a) antecedent physiologic increases of plasma cortisol can significantly blunt epinephrine, norepinephrine, glucagon, and MSNA responses to subsequent hypoglycemia and (b) these data suggest that increased plasma cortisol is the mechanism responsible for antecedent hypoglycemia causing hypoglycemia associated autonomic failure.