The objective of this study was to describe the association between the epsilon 4 allele of the apolipoprotein E gene (APOE E4) and Alzheimer's disease (AD) and to evaluate APOE E4 genotyping as a test for AD. The study base of this case-control study included about 23,000 persons 60 year of age or greater (a large health maintenance organization); the demographic characteristics of this group are similar to those of the surrounding area. Analysis focused on 234 Caucasian probable AD patients first identified between 1987 and 1993; and 304 cognitively intact controls of similar age, sex, and race who were randomly selected from the same study base. All cases were examined and diagnosed by study physicians using standard protocols. All subjects participate in continuing annual follow-up testing to verify their cognitive status. APOE genotypes were determined from blood samples using standard laboratory methods. Subject characteristics and diagnoses were obtained from interviews, diagnostic examination, or medical record review. Heterozygous E4 individuals had a crude odds ratio of 3.1 (2.1-4.5) for AD compared to those with no E4, while homozygous E4 subjects had an odds ratio of 34.3 (8.0-146.3) for AD. As an indicator of AD, having one E4 allele showed a sensitivity of 0.52 and a specificity of 0.74. Homozygous E4 genotype had a sensitivity of 0.23 and a specificity of 0.99 (when compared to non-E4 genotypes). Cardiovascular disease differed in cases and controls, but did not confound or modify the APOE E4-AI) association. In this study base, the APOE E4 allele was a significant risk factor. However, considering either homozygous or heterozygous E4 genotype as a screen or diagnostic marker for AD would miss many true cases and could misclassify many normals as AD.