Alzheimer disease (AD), the most common cause of dementia in the elderly, exists in both familial and sporadic forms. Genetic studies have led to the identification of 3 genes, beta-amyloid precursor protein (APP), presenilin-1 (PS1), and presenilin-2 (PS2), which, when mutated, can cause familial forms of AD. Mutations in each of these genes result in elevated levels of A beta42/43, a proteolytic processing fragment of APP that is deposited in brains of AD patients. Transgenic mice carrying AD-causing mutations in APP develop spontaneous age-related beta-amyloid (A beta) deposition and memory impairment. Genetic linkage and association studies have also shown that the epsilon4 allele of the apolipoprotein E (APOE) gene increases risk for AD in a dose-dependent manner in both familial and sporadic forms of AD and may account for as much as 50% of the attributable risk. APOE genotyping may be useful both as an adjunct to diagnosis and in identifying patient groups for therapeutic intervention. While the biological basis for the association of APOE epsilon4 with AD is not known, age of onset and A beta deposition are positively correlated with epsilon4 allele dosage in some cases, suggesting that this risk may also be mediated directly or indirectly through A beta. Because 50% of AD cases have no APOE epsilon4 alleles and families showing mendelian inheritance of AD exist in whom there are no mutations in any of the APP, PS1, or PS2 genes, it is likely that there are additional AD risk factors, both genetic and environmental, still to be identified.