Two-hour postprandial specimens have a -14% proportional bias for lipoprotein(a) [Lp(a)], a -0.035 g/L systematic bias for apolipoprotein (apo) A-I, and a -9% proportional bias for apo B, compared with values in 12-h fasting specimens. Although a physiological hemodilution appears to account for a portion of these biases, other major factors must be implicated for Lp(a) and apo B. Even after dilutional effects are controlled for, assayed values of Lp(a) are 11-13% lower, and assayed values of apo B are 8-9% lower, in postprandial specimens than in fasting specimens. Therefore, the time of collection of a blood sample relative to the last meal can significantly affect assayed values of lipoprotein antigens. Further studies are needed to determine whether these observations result from a physiological sequestering of lipoproteins in the postprandial state or from negative interferences affecting the assays of lipoprotein antigens.