We have investigated insulin responsiveness in relation to insulin sensitivity during sequential hyperglycemic clamping in low insulin responders (LIR), high insulin responders (HIR) and in women with a history of gestational diabetes (GD). Designation of HIR and LIR was done on the basis of mathematical modeling of the insulin response to a glucose infusion test. Insulin sensitivity was determined by a somatostatin-insulin-glucose infusion test (SIGIT) according to which LIR were subdivided into groups with higher or lesser sensitivity. Hyperglycemic clamping (60 min, 11 mmol/L of glucose) induced diphasic insulin and C-peptide responses in all groups. Insulin and C-peptide responses were significantly higher in HIR than in other groups. The ratio of first phase to total insulin response was higher in HIR v GD but did not differ between other groups. A second identical clamp was performed after a 60-minute rest period. Except in HIR, insulin levels attained were then moderately but significantly higher than during the first clamp. Conversely, the glucose utilization (mg/kg/min) to insulin (mU/L) = M/L ratio was markedly increased in LIR with high insulin sensitivity but not in other groups. We conclude that (1) large and consistent differences exist in glucose-induced insulin secretion from the pancreas between nondiabetic subjects; (2) time dynamics of insulin secretion and priming effects of glucose are similar in LIR with lesser and higher sensitivity; and (3) in the latter group a glucose stress affects insulin sensitivity more markedly than insulin responsiveness.