We investigated the association between the serum level of thrombomodulin and known coronary risk factors in 119 men who underwent coronary angiography. Total cholesterol level was significantly higher in patients with coronary atherosclerosis than in those without. Significantly higher frequency of hypertension was noted in patients with coronary atherosclerosis. Uric acid level and frequency of smoking tended to be higher in patients with coronary atherosclerosis but the differences were short short of the significant level. The serum level of thrombomodulin between patients with coronary atherosclerosis and those without was not statistically significant. Age, blood urea nitrogen, and creatinine were positively correlated and creatinine clearance was inversely correlated with the serum level of thrombomodulin. Serum levels of total cholesterol, triglyceride, high-density lipoprotein cholesterol, uric acid, and fasting blood sugar, plasma level of fibrinogen, and body mass index were not related to the serum level of thrombomodulin. There was no significant correlation between the severity of coronary atherosclerosis, hypertension, alcohol use, or smoking and the serum level of thrombomodulin. Restenosis was present in 8 of 16 patients who underwent percutaneous transluminal coronary angioplasty and had a follow-up angiogram at 6.0 +/- 3.0 months. Univariate analysis revealed no significant difference in the thrombomodulin level with and without restenosis. The present findings suggest that elevated thrombomodulin levels in patients with coronary artery disease may reflect retention of thrombomodulin due to decrease in thrombomodulin clearance in the kidney.