The authors investigated whether apolipoprotein (apo) E polymorphism has an allelic and/or genotypic impact on the risk of an ischemic vascular event (IVE) in Greek patients with cardiovascular diseases (CVD). They compared apo E polymorphisms in 1) a group of 165 patients with IVE [IVE(+)], of whom 107 had survived a myocardial infarction and 58 an ischemic stroke; 2) a group of 165 patients, matched with the first group for age and gender, with angiographically confirmed coronary artery disease but without IVE [IVE(-)]; 3) a group of 240 healthy younger individuals with no family history of CVD. The apo epsilon2 allele was 5.2-fold less frequent in the IVE(+) group compared to the IVE(-) group (1.2% vs 6.2%, p = 0.001). The frequency of the epsilon2 allele in healthy subjects was 8.1%, which is 6.7-fold higher than in the IVE(+) group (p < 0.001), and more than twice as high compared to all CVD patients (p = 0.001). No significant differences in epsilon4 allele frequencies were observed between IVE(+) and IVE(-) patients (9.8% vs 8.4%) or between patients with CVD and healthy subjects (9.1% vs 10.2%). The epsilon4 allele was not associated with an increased risk for CVD or IVE. In contrast, an inverse and beneficial association of the epsilon2 allele with IVE was observed among Greek patients with CVD. These results suggest that the epsilon4 and epsilon2 alleles have a variable significance in terms of predicting the risk of vascular events in different populations. Therefore, it is important to carry out "local" studies.