Lp(a) concentrations are largely determined by apo(a) isoform size, but several studies have shown that apo(a) isoforms could not entirely explain the increase of Lp(a) levels observed in patients with coronary heart disease (CHD). Since up to 90% of the variance in Lp(a) levels has been suggested to be attributable to the apo(a) locus, the hypothesis that polymorphisms of the apo(a) gene other than size could contribute to the increase of Lp(a) levels in CHD patients must be considered. This hypothesis was tested in the ECTIM Study comparing 594 patients with myocardial infarction and 682 control subjects in Northern Ireland and France. In addition to apo(a) phenotyping, five previously described polymorphisms of the apo(a) gene were genotyped: a (TTTTA)n repeat at position -1400 from the ATG, a G/A at -914, a C/T at -49, a G/A at -21 and a Met/Thr affecting amino acid 4168. As reported earlier [Parra HJ, Evans AE, Cambou JP, Amouyel P, Bingham A, McMaster D, Schaffer P, Douste-Blazy P, Luc G, Richard JL, Ducimetiere P, Fruchart JC, Cambien F. A case-control study of lipoprotein particles in two populations at contrasting risk for coronary heart disease. The ECTIM study. Arterioscler Thromb 1992; 12:701-707], mean Lp(a) levels were higher in cases than in controls (20.7 vs 14.6 mg/dl in Belfast, 17.2 vs 8.9 mg/dl in France, P < 0.001 for case-control and population differences). In the present study, mean apo(a) isoform size differed significantly between cases and controls (25.7 vs 26.6 kr in Belfast, 25.9 vs 27.4 kr in France, P < 0.001 for case-control and P = 0.13 for population difference). After adjustment for apo(a) isoforms, Lp(a) levels remained significantly higher in cases than in controls (difference, 4.6 mg/dl; P < 0.001). Genotype and allele frequencies did not differ significantly between cases and controls for any of the five polymorphisms studied. The five polymorphisms were in strong linkage disequilibrium and had a combined heterozygosity of 0.83. In multivariate regression analysis adjusted for apo(a) isoforms, only the (TTTTA)n polymorphism was significantly associated with Lp(a) levels; it explained 4.5% of Lp(a) variability in cases and 3.1% in controls. The Lp(a) case/control difference was not reduced after taking into account the (TTTTA)n effect. We conclude that the increase of Lp(a) levels observed in MI cases, and which was not directly attributable to apo(a) size variation, was not related to the five polymorphisms of the apo(a) gene considered.