The prevalence of abnormalities of lipoprotein cholesterol and apolipoproteins A-I and B and lipoprotein (a) [Lp(a)] was determined in 321 men (mean age 50 +/- 7 years) with angiographically documented coronary artery disease and compared with that in 901 control subjects from the Framingham Offspring Study (mean age 49 +/- 6 years) who were clinically free of coronary artery disease. After correction for sampling in hospital, beta-adrenergic medication use and effects of diet, patients had significantly higher cholesterol levels (224 +/- 53 vs. 214 +/- 36 mg/dl), triglycerides (189 +/- 95 vs. 141 +/- 104 mg/dl), low density lipoprotein (LDL) cholesterol (156 +/- 51 vs. 138 +/- 33 mg/dl), apolipoprotein B (131 +/- 37 vs. 108 +/- 33 mg/dl) and Lp(a) levels (19.9 +/- 19 vs. 14.9 +/- 17.5 mg/dl). They also had significantly lower high density lipoprotein (HDL) cholesterol (36 +/- 11 vs. 45 +/- 12 mg/dl) and apolipoprotein A-I levels (114 +/- 26 vs. 136 +/- 32 mg/dl) (all p less than 0.005). On the basis of Lipid Research Clinic 90th percentile values for triglycerides and LDL cholesterol and 10th percentile values for HDL cholesterol, the most frequent dyslipidemias were low HDL cholesterol alone (19.3% vs. 4.4%), elevated LDL cholesterol (12.1% vs. 9%), hypertriglyceridemia with low HDL cholesterol (9.7% vs. 4.2%), hypertriglyceridemia and elevated LDL cholesterol with low HDL cholesterol (3.4% vs. 0.2%) and Lp(a) excess (15.8% vs. 10%) in patients versus control subjects, respectively (p less than 0.05). Stepwise discriminant analysis indicates that smoking, hypertension, decreased apolipoprotein A-I, increased apolipoprotein B, increased Lp(a) and diabetes are all significant (p less than 0.05) factors in descending order of importance in distinguishing patients with coronary artery disease from normal control subjects. Not applying a correction for beta-adrenergic blocking agents, sampling bias and diet effects leads to a serious underestimation of the prevalence of LDL abnormalities and an overestimation of HDL abnormalities in patients with coronary artery disease. However, 35% of patients had a total cholesterol level less than 200 mg/dl after correction; of those patients, 73% had an HDL cholesterol level less than 35 mg/dl.