In massively obese patients hyperinsulinemia and insulin insensitivity usually improve with weight loss. To clarify the mechanism of these reversible abnormalities eight nondiabetic massively obese patients were studied before and at intervals (3 months and 1 yr) after weight loss following gastroplasty. Insulin dynamics were studied during the hyperglycemic clamp (change in glucose, 7 mmol L-1 for 2 h) by measuring the area under the insulin and C-peptide response curves, representing, respectively, systemic insulin response and insulin production. Compared to lean age-matched normal subjects the massively obese patients had the expected fasting hyperinsulinemia and an exaggerated insulin response (P less than 0.05). Within 3 months and after an approximately 20% weight loss, they had a marked reduction in the systemic insulin response but no change in the C-peptide response. Therefore, the reduction in insulin response was due to enhanced hepatic insulin clearance rather than reduced insulin production. Thus, the liver serves a gate-keeping role in regulating the systemic insulin response to a glucose challenge. With additional weight loss of 14% and then weight maintenance, insulin clearance was further increased, and a reduction in insulin production became evident, since the C-peptide response was reduced. Exogenous insulin clearance was measured using the euglycemic clamp technique before and after weight loss. Insulin clearance was initially lower in the massively obese patients compared to that in the normal subjects (P less than 0.05) and increased toward normal with weight loss (P less than 0.05). Similarly, insulin sensitivity, as measured by the ratio of glucose metabolised per U endogenous insulin, normalized with weight loss and weight maintenance. Thus, after significant weight loss followed by weight maintenance at a reduced, but not ideal, level, insulin clearance, production, and sensitivity all reverted to normal. These findings suggest that adipose mass per se may not be exclusively responsible for altered insulin and glucose dynamics in obesity and that additional factors associated with obesity, such as nutrient load, adipose distribution, fat cell size, or fatty acid flux, play a contributing role.