Alzheimer's disease (AD) is the most commonly diagnosed form of dementia in the elderly. Predominantly this disease is sporadic in nature with only a small percentage of patients exhibiting a familial trait. Early-onset AD may be explained by single gene defects; however, most AD cases are late onset (> 65 years) and, although there is no known definite cause for this form of the disease, there are several known risk factors. Of these, the epsilon4 allele of the apolipoprotein E (apoE) gene (APOE) is a major risk factor. The epsilon4 allele of APOE is one of three (epsilon2 epsilon3 and epsilon4) common alleles generated by cysteine/arginine substitutions at two polymorphic sites. The possession of the epsilon 4 allele is recognized as the most common identifiable genetic risk factor for late-onset AD across most populations. Unlike the pathogenic mutations in the amyloid precursor or those in the presenilins, APOE epsilon4 alleles increase the risk for AD but do not guarantee disease, even when present in homozygosity. In addition to the cysteine/arginine polymorphisms at the epsilon2/epsilon3/epsilon4 locus, polymorphisms within the proximal promoter of the APOE gene may lead to increased apoE levels by altering transcription of the APOE gene. Here we review the genetic and biochemical evidence supporting the hypothesis that regulation of apoE protein levels may contribute to the risk of AD, distinct from the well known polymorphisms at the epsilon2/epsilon3/epsilon4 locus.