The relationship of serum lipoprotein(a) [Lp(a)] to ischemic heart disease was investigated in a series of 114 men undergoing elective diagnostic coronary angiography at the University Hospital of Lausanne (Switzerland). Serum Lp(a) was higher (median: 134 mg/l; mean +/- SD: 371 +/- 509 mg/l) in the 76 individuals with ischemic heart disease, as defined by at least one significant stenosis (> or = 50%) in a proximal segment of a major coronary artery, compared to the 38 individuals without significant stenosis (60 mg/l; 155 +/- 218 mg/l). The risk ratio for ischemic heart disease was 4.2 (95% CI: 1.05-16.6) for serum Lp(a) greater than 300 mg/l compared to concentrations smaller than 50 mg/l, after adjustment for total cholesterol, HDL-cholesterol, body mass index, cigarette smoking, hypertension, and age. Lp(a) was the strongest lipid correlate of a severity score for coronary atherosclerosis (Jenkins' score) after adjustment for the other considered risk factors (partial R squared = 0.10; p < 0.001). These data are consistent with a strong independent effect of Lp(a) on ischemic heart disease. The literature on Lp(a) is reviewed and a clinical approach proposed.