Epidemiological data suggest that a prolonged difference of 0.6 mmol.L-1 (approximately 10%) in total cholesterol results in an average reduction in relative risk of coronary heart disease (CHD) of 27% in cohort studies and 38% in international studies. Five large, randomized controlled trials of lipid reduction by the HMG-CoA– reductase inhibitors (statins) have shown unequivocal clinical benefit in primary and secondary prevention. The crude risk reduction per 0.6 mmol.L-1 cholesterol in these trials was 12% to 21% (average 15%), approximately 50% of the epidemiologically predicted effect. If allowance is made for the impact of regression dilution bias, for possible overestimation of the association between cholesterol levels and CHD risk in epidemiological studies, for poor or noncompliance, and for the limited duration of the trials, it is likely that the true benefit of long-term statin therapy will be close to that predicted from cohort studies...