A positive family history is associated with increased risk for dementia. It is not known whether a negative family history with long-lived relatives predicts a reduced risk for dementia. We studied the survival rate and the occurrence of dementia in 232 parents and siblings of 43 optimally healthy individuals ≥84 years of age and compared them with 233 parents and siblings of 51 random controls and 499 parents and siblings of 88 Alzheimer disease (AD) patients. Prevalence of dementia after age 60 years was .031 for the relatives of healthy elderly, .066 for the relatives of random controls, and .217 for the relatives of AD patients. The cumulative incidence of dementia by age 85 years was estimated as .041 (±.019) for the relatives of healthy elderly individuals, .102 (±.038) for the relatives of random controls, and .360 (±.037) for the relatives of AD patients. Hazard-ratio estimates suggest that the risk of dementia for the relatives of healthy elderly is 3 times lower than the risk for the relatives of random controls (P < .03) and is 11 times lower than the risk for the relatives of AD patients (P < .00005). An analysis of age at death indicated that the relatives of healthy elderly and the relatives of AD patients had a longer life span than did the relatives of random controls. These results suggest (1) that a segment of the population is at considerably lower risk for dementia and that a documented negative family history can help identify these individuals, (2) that, for genetic counseling, it is important to document a negative family history of dementia as well as a positive family history, and (3) that familial/genetic factors exist that reduce the risk of dementia and that association studies utilizing optimally healthy elderly may help identify genes that promote successful aging of the brain.