Analysis of epidemiological studies enables definition of the optimal serum cholesterol level between 2 and 2.2 g/l (5.2 and 5.7 mmol/l). Higher levels are associated with an exponential risk of coronary artery disease: lower levels should be interpreted with caution because of the J-shaped curve of global mortality reported in several trials. In primary prevention, therapeutic trials have clearly shown that a significant reduction in the number of coronary event may be obtained by lowering the serum cholesterol (15 for every 1,000 patients treated). In trials performed to date this benefit has not resulted in any gain in global mortality. In secondary prevention, the benefits of lowering serum cholesterol on the incidence of coronary disease seems identical in terms of relative risk to that observed in primary prevention. Nevertheless, the benefits in terms of absolute risk are much higher because of the high incidence of coronary mortality in a patient population with a previous coronary event. Thus, for 1,000 patients treated, there are 50 less cardiac events and this is reflected in a reduction of global mortality of about 20 for every 1,000 patients treated.