The current prevalent view is that plasma lipoprotein(a) [Lp(a)] concentrations are under strong genetic control. Most dietary and drug interventions seem to have little or no effect on plasma Lp(a) levels. However, evidence for a possible regulatory rol e of hormones is accumulating, for instance, fluctuations of Lp(a) levels during pregnancy have been reported. Also, in insulin-dependent diabetes mellitus (IDDM) patients, elevated Lp(a) levels have been reported. In the present longitudinal study, plasma lipid concentrations, including Lp(a), were determined in IDDM women before pregnancy, during pregnancy, and 3 months postpartum. In our study population, Lp(a) concentration was not significantly correlated with either hemoglobin A1c (HbA1c) levels of apolipoprotein(a) [apo(a)] phenotype. Changes in other lipid parameters observed during pregnancy in our IDDM population were similar to those reported during normal pregnancy. Lp(a) concentrations were quantified using two different immunochemical methods that possess different sensitivities and specificities: an immunoradiometric assay (IRMA) using two different anti-apo(a) antibodies, and an enzyme-linked immunosorbent assay (ELISA) using an anti-apo(a) and an anti-apo B antibody. Median prepregnancy Lp(a) concentrations were 118 mg/L (range, 15 to 672) as determined with the IRMA and 107 mg/L (range, 21 to 451) as determined with the ELISA. Women with IDDM showed, in general, no significant change in Lp(a) concentration during pregnancy when it was assayed with the IRMA, although a tendency to increased values was observed. When Lp(a) concentrations were determined with the ELISA, a strong and significant increase in Lp(a) from weeks 17 to 24 of pregnancy onward was found. The latter results confirm the prevalent view that during pregnancy Lp(a) levels are increased. However, the present results and those of others and Lp(a) in normal pregnancy strongly emphasize the importance of method selection when determining Lp(a) concentrations.