Despite similar glycemic profiles, higher insulin levels are achieved following oral versus intravenous administration of glucose. This discrepancy is due to the incretin effect and is believed to be mediated via stimulation of beta-cells by hormone(s) released from the gut. The leading gut hormone candidates are glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide (GLP-1). To determine the relative insulinotropic activity of these peptides, we infused GLP-1(7-37) and GIP into normal subjects and patients with non-insulin dependent diabetes mellitus (NIDDM). In normal subjects during euglycemia, GLP-1(7-37) stimulated insulin release, whereas GIP did not. Using the Andres clamp technique, we established stable hyperglycemia for 2 h (5.4 mmol/l above the basal level). During the second hour, either GIP, GLP-1(7-37), or both were infused in normal healthy volunteers and in patients with NIDDM. In normal subjects, at a glucose level of 10.4 mmol/l, the 90-120 min insulin response was 279 pmol/l. GIP at a dose of 1, 2 or 4 pmol/kg/min augmented the 90-120 min insulin response by 69, 841 and 920 pmol/l, while GLP-1(7-37), at a dose of 1.5 pmol/kg/min augmented the insulin response by 2106 pmol/l. When both hormones were administered simultaneously, the augmentation was additive--2813 pmol/l. In the diabetic subjects, GIP had no effect, while GLP-1(7-37) augmented the insulin response by 929 pmol/l. We conclude that in normal healthy subjects, GLP-1(7-37), on a molar basis, is several times more potent than GIP at equivalent glycemic conditions. The additive insulinotropic effect suggests that more than one incretin may be responsible for the greater insulin levels observed following oral administration of glucose compared to the intravenous route. In NIDDM, GIP had no insulinotropic effect, while GLP-1(7-37) had a marked effect. This suggests that GLP-1(7-37) may have therapeutic potential as a hypoglycemic agent in NIDDM patients.