Gastric inhibitory polypeptide, or GIP, has been postulated as the major enteric hormonal mediator of insulin release. The release of immuno-reactive GIP (IR-GIP) after oral glucose and its role in insulin release was studied in normal men by the glucose clamp technique. In 24 subjects studied with the hyperglycemic clamp, blood glucose was maintained at 125 mg/dl above basal for 2 h via a primed-continuous IV glucose infusion coupled to a servo-controlled negative feedback system. 40 g glucose per m(2) surface area was ingested at 60 min, and the blood glucose was maintained at the steady-state hyperglycemic level. Plasma IR-GIP and insulin (IRI) levels were measured throughout the 2-h period. IR-GIP levels changed little when IV glucose alone was given; the mean basal value was 305+/-34 (SEM) pg/ml. After oral glucose, IR-GIP levels began to rise within 10 min and reached a peak within 40 min of 752+/-105 pg/ml. Plasma IRI responded initially to the square wave of hyperglycemia in the typical biphasic pattern. After oral glucose, plasma IRI levels rose strikingly above the elevated levels produced by hyperglycemia alone, reaching a peak of 170+/-15 muU/ml within 45 min. The time course of the rise in IR-GIP and IRI was nearly identical. To assess whether the maintenance of euglycemia would affect this process, the euglycemic clamp was employed in 11 subjects to maintain basal blood glucose levels during a similar 2-h study. A primed-continuous insulin infusion, with a constant rate of 120 mU/m(2) per min was given together with a servo-controlled glucose infusion. This resulted in hyper-insulinemia of approximately 300 muU/ml. Glucose was ingested by six subjects at 60 min. Plasma IR-GIP responded to oral glucose similarly to the effect seen in the hyperglycemic studies. No increase in endogenous insulin release was seen despite the increase in IR-GIP when euglycemia was maintained. However, in five of seven subjects given insulin whose blood glucose concentration rose by 20 mg/dl or more after oral glucose, there was an increase in plasma insulin concentration associated with the elevation in IR-GIP. Thus, the effect of glucose-released IR-GIP on insulin secretion is dependent upon the presence of some degree of hyper-glycemia and is not inhibited in the presence of marked hyperinsulinemia.