Since the first description of Alzheimer's disease (AD) at the beginning of the century until relatively recently, it was customary to define Alzheimer's disease as occurring in the presenium. The same neuropathological changes occurring in brains over the age of 65 were called "senile dementia." Because there have been no clinical or pathological features to separate the two groups, this somewhat arbitrary distinction has been abandoned. Although AD is currently considered to be a heterogeneous disease, the most consistent risk factor to be implicated other than advancing age is the presence of a positive family history. This potential genetic vulnerability to AD has been recognized for some time. Some of the earliest evidence suggestive of a genetic contribution to AD came from Kallmann's 1956 study (1) demonstrating a higher concordance rate in monozygotic twins for "parenchymatous senile dementia" compared with dizygotic twins and siblings. This monozygotic excess has been confirmed in studies applying more rigorous diagnostic criteria although there may be widely disparate ages of onset between twins (2). The most convincing evidence for a genetic contribution to AD has come form the study of pedigrees in which the pattern of disease segregation can be clearly defined. Thus, the abandonment of the early and late-onset dichotomy has occurred at a time when, at the genetic level, important differences have been identified through the discovery of specific gene defects in early onset cases.