Abstract Five different gel systems were evaluated for their utility in determining the molecular weights of apolipoprotein(a) (apo(a)) polymorphs by SDS polyacrylamide or agarose gel electrophoresis. Three linear polyacrylamide gradient gels (2–16% from Isolab (Akron, OH), 4–15% from Pharmacia (Piscataway, NJ), and 2.5–6% homemade), a 4% polyacrylamide, and a 1.5% agarose gel were examined. Crosslinked phosphorylase B oligomers served as molecular weight standards. Molecular weights of four different apo(a) polymorphs were determined in each gel system and compared to values measured previously by sedimentation equilibrium. The results indicate that molecular weights obtained by gradient polyacrylamide gel electrophoresis were within 10% and often not statistically different from values acquired by sedimentation equilibrium. The use of homogenous 4% polyacrylamide and 1.5% agarose gels led to molecular weights that were overestimated by 20 and 60–70%, respectively. ApoB 100, which is a commonly used molecular weight marker, was found to have anomalously fast mobility in each of the four polyacrylamide gel systems. Because its use would lead to overestimated apo(a) molecular weights, it was not useful as a molecular weight standard. Our results indicate that SDS–gradient polyacrylamide gel electrophoresis with cross-linked phosphorylase B as standard is a suitable gel system for evaluating apo(a) molecular weights.