To determine whether waist/height ratio is a useful predictor for coronary heart disease (CHD) risk factors in men. An epidemiologic study comparing relationships between waist/height ratio, body mass index, waist/hip ratio and CHD risk factor levels [continuous variables--systolic and dialostic blood pressure, fasting blood glucose, hemoglobin A1c (HbA1c), triglyceride, cholesterol, HDL cholesterol values] and the risk factor morbidity index (sum of the risk factor scores for hypertension, abnormal glucose tolerance, hypertriglyceridemia, hypercholesterolemia and low HDL cholesterol--one point each if present). A health examination facility within a general hospital. 3131 men underwent routine health examination. Body mass index, waist/hip ratio and waist/height ratio were significantly associated with all of the risk factor levels and with the risk factor morbidity index according to the result of simple regression analysis. Multiple regression analysis for waist/hip ratio and body mass index showed that both of them were also significantly associated with all of the risk factor levels and with the risk factor morbidity index, except HbA1c levels with body mass index. According to the results of multiple regression analysis for waist/height ratio and body mass index, body mass index was not significantly associated with fasting blood glucose, HbA1c, cholesterol levels or the risk factor morbidity index whereas waist/height ratio was significantly associated with all of the variables. On the other hand, while waist/height ratio was significantly associated with all of the variables, waist/hip ratio was not significantly associated with most of the variables when waist/height ratio and waist/hip ratio were compared. Waist/height ratio, an index of abdominal obesity, may be a better predictor of multiple CHD risk factors in men than waist/hip ratio in mass epidemiologic studies.