To define the pathogenic factors responsible for glucose intolerance in NIDDM, we estimated insulin secretory capacity, SI, and SG in 11 healthy, nondiabetic subjects and 9 NIDDM patients who had no SI impairment. All subjects studied were nonobese and normotensive. Each underwent a 75-g OGTT and a modified FSIGT: glucose was administered (300 mg/kg body weight), and insulin was infused (20 mU/kg over 5 min) from 20 to 25 min after the administration of glucose. SI and SG were estimated by Bergman's minimal-model method. The insulin response to oral glucose was significantly lower in NIDDM patients than in normal control subjects. First-phase insulin secretion expressed as the integrated area of plasma insulin above the basal level during the first 20 min was much smaller in NIDDM subjects (214 +/- 112 pM.min) than in control subjects (4643 +/- 885 pM.min, P < 0.01). SI was not statistically different in normal control subjects (1.27 +/- 0.18 x 10(-4) min-1.pM-1) versus diabetic patients (1.62 +/- 0.33 x 10(-4) min-1.pM-1). However, SG was significantly lower in diabetic subjects (1.11 +/- 0.17 x 10(-2) min-1) than in control subjects (2.35 +/- 0.26 x 10(-2) min-1, P < 0.01). These results suggest that impaired insulin secretion and decreased SG are the factors responsible for glucose intolerance of Japanese NIDDM patients with normal insulin sensitivity. Because SI and SG are the factors responsible for glucose intolerance of NIDDM patients with insulin resistance, it is conceivable that decreased SG is common in NIDDM patients regardless of their SI index.