Late-onset Alzheimer’s dementia (LOAD) is terminal and the most prevalent form of dementia (70% of cases). LOAD is a major public health burden in the US with current prevalence estimates of 4 to 5 million adults and economic costs exceeding $236 billion annually. Important disparities in LOAD prevalence occur with sex, race/ethnicity, education, and residence. The biological bases of these health disparities are incompletely characterized and their influences on LOAD are likely to be multifactorial. Thus, studies with sufficient sample sizes, concurrently assessing multiple characteristics, such as educational attainment, environment, social, behavioral, lifestyle, geographic, and genetics, will be uniquely positioned to effectively test factors or combinations of factors that create and sustain LOAD disparities. Our goal is to determine the joint genetic and environmental contributions to LOAD risk that underlie these health disparities. Using existing genomics data, well-characterized dementia phenotypes, and diverse risk factor data, we will analyze up to 16,000 aging participants in the Health and Retirement Study (HRS), attempt clinical confirmation in participants of the Aging, Demographics, and Memory Study (ADAMS) sub-cohort of the HRS, and test replication in other clinical and population-based samples. Our aims are to (1) determine the cumulative genetic risk of LOAD by testing the association between cognitive polygenic scores and risk of dementia phenotype in European and African ancestries; (2) determine the polygenic effect of LOAD risk covariates from behavioral, physiological, and psychosocial domains on dementia phenotypes in European and African ancestries; and (3) test for effect modification of the association between polygenic risk and dementia phenotype in European and African ancestries by health disparities factors (sex, education, urban/rural residence). This study will likely impact the field of Alzheimer’s research and contribute to public health because it will a) establish the relevance of cumulative genetic risk on LOAD in susceptible populations where genetics may be a more relevant factor; b) elucidate important biological mechanisms through polygenic scores; c) determine the combined and individual gene-environment contribution to LOAD risk; d) generate unified polygenic scores and dementia phenotypes across major longitudinal cohorts involving diverse populations that can be used in future investigations of health outcomes and/or additional exposure domains; e) consider the effects of sex, educational attainment, ancestry, and urban/rural status in the same study where comparisons of relative contribution to risk can be made. We have the opportunity to simultaneously and significantly improve our understanding of the genetic and environmental etiologic contributions to health disparities in LOAD.