AIM To find out from published reports whether the tendency of coronary heart disease (CHD) to progress can be retarded by lowering total and LDL cholesterol concentrations. METHODS After a data-base search a meta-analysis was undertaken of all those randomized, controlled and angiographically documented studies which contained informations about the effect of cholesterol reduction on the course of CHD over a period of at least 2 years. A total of 12 studies covering 3781 patients met the stated criteria. RESULTS The different lipid-lowering measures (usually drugs) achieved a statistically significant reduction of the number of patients with progression of the coronary angiographic findings and a significant increase in the number of those with actual regression. The number of coronary incidents, such as myocardial infarction, unstable angina, sudden cardiac death, necessary aortocoronary bypass operations or percutaneous transluminal coronary angioplasty was lower by 34% in the groups with measures to influence cholesterol metabolism than in the untreated groups. In the studies in which coronary arteriograms were evaluated quantitatively there was an annual increase in the mean degree of stenosis of 1.01% in the untreated and 0.37% in the treated groups. CONCLUSIONS The difference in the increase of the mean degree of stenosis is probably not only important because of the resulting haemodynamic changes in myocardial perfusion. The quantitatively only slightly increased trend towards progression in the untreated groups can be interpreted as a marker for an active process of atherogenesis with a tendency of plaques to break off and of plaque thromboses.