The discovery of pathogenic mutations in the amyloid precursor protein (APP) gene and the presenilin (PS1, PS2) genes, causing familial early-onset AD has lead to the hypothesis of the amyloid cascade. The epsilon4 allele of the apolipoprotein E (APOE) gene, the only recognized genetic risk factor for AD, may be involved in the mechanism. However, to date, search for new genetic determinants has been hampered by methodological limitations. Some loci, for instance on chromosome 12, have been characterized by linkage studies performed in familial cases, but the regions of interest are very large and contain numerous genes. Furthermore, search for polymorphisms implicated in the development of AD, should not be limited to the coding part of the genes, but should also involve the non-translated sequences of the genes, for instance in the regions regulating gene expression. Indeed, these genetic variations may have important impact on key proteins of the pathologic process. Although this task is difficult, the identification of new susceptibility genes should lead to a better understanding of the development of AD.