Using the resources of the Rochester Epidemiology Project, we conducted a case-control study of sociodemographic characteristics using the incidence cohort of patients with Alzheimer's disease (AD). During the conduct of ongoing studies of AD in Rochester, we identified new cases of AD as they occurred during 1975-1979 (N = 241). We selected one age- and sex-matched control from among all registrations for care in this community during the year of onset for each case. There was little difference between cases and controls for educational attainment, marital status, type of dwelling, living arrangement, or occupation. We were unable to confirm low educational level as a risk factor for AD in this population. Future attempts to identify etiologic risk factors for dementing illness should probably move toward other areas of research.