Plasma lipids, lipoproteins and apolipoproteins A1 and B were determined in a reference population and in 273 patients who underwent coronary arteriography for evaluation of chest pain. The patients were divided into 3 groups according to arteriographic findings. The 102 patients in group A had normal coronary arteries; the 30 patients in group B had small vascular abnormalities but no significant narrowing; the 141 patients in group C had significant vascular lesions with greater than or equal to 50% narrowing. Total cholesterol and triglycerides were quantified using enzymatic methods. HDL cholesterol and phospholipids were enzymatically analyzed after phosphotungstate precipitation. Plasma lipoproteins were evaluated by quantitative electrophoresis based on densitometric scanning of the lipoprotein bands after separation and visualization by polyanion precipitation. Apoproteins A1 and B were measured by electroimmuno-diffusion assay using ready-for-use plates. Statistical evaluation of the data showed that group C patients had higher betalipoprotein and apo B levels and lower alphalipoprotein and apo A1 levels than subjects with normal coronary arteries. The predictive value of these results was established in relation to clinical findings. Plasma lipid measurements proved inadequate to separate the 3 groups. In contrast, the betalipoprotein/alphalipoprotein ratio and, still better, the apo A1/apo B ratio provided good discrimination. The results suggest that measurement of plasma apo A1 and apo B levels is the best means of assessing the risk of coronary artery disease.