AbstractThe demand for high density lipoprotein (HDL) quantitation has dramatically increased with the renewed awareness of the importance of HDL as a negative risk factor for coronary heart disease. HDL is usually estimated by specific precipitation of the non‐HDL apoB‐containing lipoproteins by polyanions and divalent cations followed by measurement of cholesterol in the supernatant. A common procedure involves precipitation with sodium heparin at 1.3 mg/ml and MnCl2 at 0.046 M (final concentrations). This method is appropriate for serum but less than ideal for plasma because of incomplete precipitation and sedimentation of the apoB‐containing lipoproteins. A two‐fold increase in Mn2+ to 0.096 M improves precipitation of the apoB‐associated lipoproteins from plasma without excessive precipitation of HDL. This modified heparin‐Mn2+ procedure gives results nearly identical to the results with the ultracentrifugal reference method (cholesterol in the d>1.063 fraction corrected for losses and the presence of apoB‐associated cholesterol). The dextran sulfate 500‐Mg2+ and the sodium phosphotungstate‐Mg2+ procedures give results consistently 2–4 mg/dl lower than does the reference method. In contrast, a heparin‐Ca2+ method gives results 5–8 mg/dl higher than does the reference method. Immunochemical analysis of apoA‐I in the precipitate and apoB in the supernatant indicates that lower values for the phosphotungstate‐Mg2+ procedure is due to partial precipitation of the A‐I‐containing lipoproteins, while higher values by the heparin‐Ca2+ method are due to incomplete precipitation of the apoB‐containing lipoproteins. Quantitation of the principal apoproteins of HDL, A‐I and A‐II, represent an important additional index of HDL concentrations and composition. Quantitation of plasma A‐I and A‐II concentrations by radial immunodiffusion indicates that women generally have higher HDL concentrations than men (women, A‐I, 135±25, A‐II, 36±6; men, A‐I, 120±20, A‐II, 33±5; mean±S.D., in mg/dl). A‐I and A‐II do not increase with age in men but show a slight increase with age in women. Estrogen increases HDL cholesterol and protein and may in part account for the higher HDL in women. The lighter density HDL subclass has a higher A‐I/A‐II ratio than the denser HDL subclass, with women generally having significantly more of the lighter HDL subclass. Density‐gradient ultracentrifugation in CsCl2 gradients indicates that HDL contains subpopulations of differing hydrated density which vary in the A‐I/A=II ratio. Immunoassay of A‐I and A‐II when used in combination with HDL cholesterol analysis is a powerful tool for studies of HDL structure, epidemiology and metabolism.