Subjects with impaired glucose tolerance (IGT) are usually overweight and exhibit insulin resistance with a defective compensation of insulin secretion. In this study, we sought to establish the interrelation between insulin secretion and insulin sensitivity after oral glucose in non-obese subjects with IGT and we also examined this interrelation in relation to the 2 main incretins, glucagon-like peptide (GLP-1) and gastric inhibitory polypeptide (GIP). To that end, 13 women with IGT and 17 women with normal glucose tolerance (NGT) underwent an oral glucose tolerance test (OGTT) with measurements of glucose, insulin, C-peptide, GLP-1, and GIP. Insulin secretion (TIS) and insulin sensitivity (OGIS) were assessed using models describing the relationship between glucose, insulin and C-peptide data. These models allowed estimation also of the hepatic extraction of insulin. The age (54.2 +/- 9.7 [mean +/- SD] years) and body mass index (BMI; 26.0 +/- 4.0 kg/m(2)) did not differ between the groups. Subjects with IGT displayed lower TIS during the initial 30 minutes after oral glucose (0.97 +/- 0.17 [mean +/- SEM] v 1.75 +/- 0.23 nmol/L in NGT; P =.018) and lower OGIS (397 +/- 21 v 463 +/- 12 mL/min/m(2); P =.005). The incremental 30-minute TIS times OGIS (reflecting insulin secretion in relation to insulin sensitivity) was significantly reduced in IGT (359 +/- 51 v 774 +/- 91 nmol/min/m(2), P =.001). This measure correlated inversely to the 2-hour glucose level (r = -0.71; P <.001). In contrast, TIS over the whole 180-minute period was higher in IGT (26.2 +/- 2.4 v 20.0 +/- 2.0 nmol/L; P =.035). Hepatic insulin extraction correlated linearly with OGIS (r = 0.71; P <.001), but was not significantly different between the groups although there was a trend with lower extraction in IGT (P =.055). Plasma levels of GLP-1 and GIP increased after oral glucose. Total secretion of these incretin hormones during the 3-hour test did not differ between the 2 groups. However, the 30-minute increase in GLP-1 concentrations was lower in IGT than in NGT (P =.036). We conclude that also in non-obese subjects with IGT, when adiposity is controlled for in relation to NGT, defective early insulin secretion after oral glucose is a key factor. This defective beta-cell function is associated with, and may be caused by, a reduced early GLP-1 response.