GWAS studies were very successful in identifying genetic loci associated with AD risk. However, these studies could not point to the actual causal variant. In this study, WES and/or WGS data is being generated from families densely affected by Alzheimer disease (AD) under the hypothesis that these families will be enriched for rare genetic variants that confer risk to AD.  This study included late onset families with a high proportion of AD cases (at least three affected members with DNA data available). The families originated from the NCRAD, NIA-LOAD, Knight-ADRC and DIAN-EXR cohorts. The proband of the family could not be a carrier of known pathogenic mutations in the Mendelian genes for AD (APP, PSEN1, PSEN2) or Frontotemporal Dementia (GRN, MAPT, C9ORF72).  Cases within families had to be 65 years old or older at age at onset (AAO) and have a CDR >=0.5; controls within families had to have a CDR=0 at last assessment and be older than the largest AAO within the family.  AD definition was based on a combination of both clinical and pathological information if available. Clinical diagnosis was overruled by pathological diagnosis.  Subjects are Non-Hispanic white from North America, based on self-reporting. Autopsy information was provided if available, but it was not a requirement for enrollment.