While lowering the blood cholesterol level has been shown to decrease the risk of coronary heart disease (CHD) in Western countries, little is known about the relationship between lipid-lowering drugs and CHD risk in Japan. A case - control study of nonfatal acute myocardial infarction (AMI) was conducted in Japan. A total of 658 cases aged 40-79 years and 1,274 community controls matched to each case by sex, year of birth, and residence were compared for their use of lipid-lowering drugs and level of serum total cholesterol. The study used 2 definitions of hypercholesterolemia: (i) either serum total cholesterol ≥6.21 mmol/L (240 mg/dl) or current use of hypolipidemic drugs; and (ii) either serum total cholesterol of ≥5.69 mmol/L (220 mg/dl) or current use of hypolipidemic drugs. Unconditional logistic regression analysis was used to adjust for hypertension, angina pectoris, diabetes mellitus, smoking, alcohol use, and physical activity as well as for sex and age. As compared with nonhypercholesterolemic subjects, those with untreated hypercholesterolemia had a significantly increased risk of AMI whereas no increase in the risk was observed for treated hypercholesterolemia. In the group with hypercholesterolemia, regardless of the definition, the adjusted odds ratios decreased progressively when the achieved cholesterol levels were lower. (Circ J 2002; 66: 463 - 468)