The aim of this study was to clarify whether insulin resistance contributes to atherosclerosis in patients with non-insulin-dependent diabetes mellitus (NIDDM). Fifty-three NIDDM patients (36 males and 17 females, 53+/-10 years old (mean+/-SD)) were studied. As an index of atherosclerosis, we measured the average thickness (IMT) as well as basal thickness excluding the maximum thickness and the height of the maximum thickness of the carotid artery wall. Euglycemic hyperinsulinemic glucose clamp was conducted for 90 min to evaluate average glucose infusion rate (GIR) as an index of insulin sensitivity in the peripheral tissues. For another 180 min after intake of oral glucose load with 0.3 g/kg, the euglycemic hyperinsulinemic clamp was continued to measure ratio of splanchnic glucose uptake (SGU) as an index of insulin sensitivity of the liver. The patients were separated into three activity groups according to the grade of their leisure-time physical activity. GIR (r = -0.32, p < 0.05) but not SGU (r=0.139) showed a significant inverse relationship with IMT. Multivariant regression analysis indicated that age and total cholesterol remain as independent risk factors for basal thickness and GIR as only independent risk factor for the height of the maximum thickness. Paralleling the degrees of habitual exercise (low, moderate, and high active group), GIR was higher (6.19+/-1.02, 6.38+/-1.38, 7.44+/-1.80, respectively) and IMT was lower (1.34+/-0.33 mm, 1.20+/-0.31 mm, and 1.12+/-0.29 mm, respectively) in male NIDDM as well as in female NIDDM. These data suggest that insulin resistance in the peripheral tissues but not the splanchnic tissues may independently contribute to carotid arterial wall thickness and especially to plaque lesion, and that habitual exercise might reduce insulin resistance leading to attenuation of atherosclerosis.