Several studies have shown an association between cardiovascular disease and high Lp(a) levels. Lp(a) levels are genetically determined in normal subjects. The genetic origin of elevated Lp(a) levels in patients with myocardial infarction (MI) requires confirmation however, as the results of studies involving the determination of apo(a) phenotypes in such populations are not entirely consistent. The distribution of apo(a) isoforms and of 5 different polymorphisms of the apo(a) gene, including a (TTTTA)n repeat at -1400, a G/A at -914, a G/A at -49, a C/T at -21, and a Met/Thr at position 66 of KrIV-10, have been determined in the ECTIM population. ECTIM is a case/control study comprising MI patients (n = 639) and controls (n = 761), from Northern Ireland and France. As previously reported, mean Lp(a) levels were significantly higher in patients than in controls (18.8 24.9 vs 11.1 16.9 mg/dl). Mean apo(a) isoform size was also significantly different between patients and controls (25.5 5.1 vs 26.6 5.2kr, in Belfast and 26.1 4.9 vs 27.4 4.7kr, in France), but explained only 40 of the case/control difference in Lp(a) levels. In contrast, the five polymorphisms of the apo(a) gene did not differ significantly between patients and controls and explained only a small fraction of Lp(a) variance (<4). Nonetheless, they were highly informative, as attested by their high combined heterozygosity (0.83). We conclude that the increase of Lp(a) levels observed in patients with MI, and which was not directly attributable to apo(a) size variation, is not likely to be related to further variations in the apo(a) gene. Sources of variation other than apo(a) size, which could be either genetic, environmental or metabolic in nature, contribute to the high Lp(a) levels observed in MI patients