Many studies have examined the role of APOE genotype in the development of dementia, specifically Alzheimer disease (AD). The APOE epsilon4 allele (APOE4) is a risk factor for both clinical and neuropathologic AD whereas the APOE epsilon2 allele (APOE2) seems to be protective. This would predict, even with advanced age, that APOE2 carriers would be less likely to have dementia and less likely to meet pathologic criteria for AD.The first 85 genotyped participants from The 90+ Study to come to autopsy were included. All-cause dementia (using DSM-IV criteria) and AD (using National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association criteria) diagnoses were made by consensus conference using all available information including neuropsychological testing, neurologic examination, and medical records. Neuropathologic examination included Braak and Braak staging for plaques and tangles and diagnosis of neuropathologic AD using National Institute on Aging-Reagan criteria.Across all genotypes, 58.5% of subjects were diagnosed with clinical dementia (81% of dementia was AD) and 50.0% met neuropathologic criteria for AD. Compared to those with an APOE epsilon3/epsilon3 genotype (APOE3/3), APOE4 carriers were more likely to be diagnosed with dementia (odds ratio [OR] = 12.2, 95% confidence interval [CI] = 1.5-102.0), whereas APOE2 carriers were not (OR = 0.3, 95% CI = 0.1-1.3). Surprisingly, both APOE4 (OR = 4.6, 95% CI = 1.3-16.5) and APOE2 (OR = 7.8, 95% CI = 1.5-40.2) carriers were more likely to meet neuropathologic criteria for AD than those with APOE3/3 genotype.In the oldest old, the presence of the APOE epsilon2 allele (APOE2) was associated with a somewhat reduced risk of dementia, but paradoxically was associated with increased Alzheimer disease (AD) neuropathology. Therefore, oldest old APOE2 carriers may have some mechanism that contributes to the maintenance of cognition independently of the formation of AD pathology.