BACKGROUND Apolipoproteins have been recently suggested as an alternative to lipoproteins in prediction of cardiovascular risk. Data regarding their added value in predicting the prognosis of coronary heart disease (CHD) patients are scarce. Our aim was to examine the association between lipoprotein cholesterol and related apolipoproteins with long-term mortality among CHD patients. METHODS Patients (4,472 men; 624 women, 40-74 years old) with total cholesterol <270 mg/dL (<7.0 mmol/L), high-density lipoprotein cholesterol (HDL-C) <45 mg/dL (<1.16 mmol/L), and triglycerides <300 mg/dL (< or =3.39 mmol/L); excluded from the Bezafibrate Infarction Prevention study; or included in the placebo arm were followed up for a median of 12.3 years. RESULTS Among both men and women, the association of apolipoproteins A-I and B with mortality was comparable to their corresponding lipids (HDL-C, non-HDL-C respectively). Adjusting for age, disease history, comorbidities, smoking and baseline glucose, the risk associated with the upper versus the lower tertile (lower vs upper for HDL-C and apolipoprotein A-I) among men were 1.04 (95% CI 0.91-1.19) for non-HDL-C; 1.11 (0.97-1.27) for apolipoprotein B; 1.24 (1.09-1.41) for HDL-C; and 1.30 (1.14-1.49) for apolipoprotein A-I. Atherogenic to nonatherogenic particle ratios (lipids or apolipoproteins) were in line with the results of their individual components pointing to a less atherogenic profile among women. Models including either apolipoprotein or cholesterol subfractions had similar predictive power. CONCLUSIONS Lipoprotein cholesterol and associated apolipoprotein have comparable ability to predict long-term mortality. The measurement of apolipoproteins constitutes an acceptable alternative to the use of blood lipids in assessing prognosis for CHD patients.