Elevated plasma or serum lipoprotein(a) (Lp(a)) levels have been associated with premature coronary heart disease (CHD). Lp(a) levels can be assessed quantitatively by electrophoresis and quantitatively by immunoassays determining either total Lp(a) mass, apo(a) mass on Lp(a) protein mass, or by precipitation methods followed by measurement of Lp(a) cholesterol. We prefer the latter method because it can be standardized. Electrophoretic methods can detect total Lp(a) values > or = 30 mg/dl. These values correspond to Lp(a) cholesterol values > or = 10 mg/dl. Such values are above the 75th percentile and represent high risk values for CHD. Values above the 90th percentile for middle aged men and women in Framingham (n = 2678) are > or = 38 mg/dl for total Lp(a). About 16% of patients with premature CHD (n = 321) have such values and have familial Lp(a) excess. Lp(a) is atherogenic because it can be deposited in the arterial wall, and it also can interfere with fibrinolysis. Multiple apo(a) isoforms have been found and are due to a variable number of kringle 4 like repeats. Lower molecular weight apo(a) isoforms forms are associated with elevated Lp(a) values and are more frequent in CHD kindreds. Both Lp(a) levels and apo(a) isoforms are highly heritable in this Caucasian population. Lp(a) values can be decreased with niacin, and such therapy should be strongly considered in CHD patients with elevated Lp(a) levels (> or = 30 mg/dl) since niacin treatment has been shown to decrease CHD morbidity and mortality in unselected CHD patients.