For a long time, many clinicians assumed that as an aging-related disease, Alzheimer disease (AD) was inevitable at extreme old age. In reaction in part to some earlier prevalence studies contending the same, John Morris1 contended otherwise. First, Morris pointed out a number of common mistakes that likely lead to the supposition of inevitability. Then, he went on to describe longitudinal comprehensive neuropsychological testing results correlated with quantitative postmortem studies that indicated there can be a dissociation between AD and aging and that there are people who remain cognitively intact and without neuropathological evidence well into their nineties. Morris warned that we should keep the following in mind when undertaking epidemiological study of AD: (1) cross-sectional studies in which 1 age cohort is compared against another are highly prone to confounding, for instance by years of education and other environmental exposures; (2) cross-sectional studies are less reliable in establishing a neurocognitive diagnosis, particularly at the early stages of disease; (3) multiple and reliable cognitive tests are required to test different cognitive function domains in order to achieve acceptable sensitivity for early impairment; (4) rather than subscribing to the ageist generalization that a person is doing well given their extreme age, we should rely on sensitive and specific diagnostic criteria no matter what the age; and (5) longitudinal study is far superior to cross-sectional study for achieving reliable levels of sensitivity and specificity. In my own experience as a geriatrics fellow, having 2 centenarian patients who appeared to be completely cognitively intact was my first tipoff that at least some extremely old individuals must have a resistance to AD.