This paper reexamines recent epidemiologic and molecular genetic studies on the genetic basis of Alzheimer's Disease (AD). Careful analysis of the available epidemiologic data strongly suggests that at least a proportion of AD results from the inheritance of an autosomal dominant gene defect. However, studies of isolated families, of concordance rates in twins, and of risk for AD in relatives of AD probands yield conflicting data. While it is likely that much of the conflict can be ascribed to methodologic differences, it remains premature to conclude that all AD is transmitted as an autosomal dominant trait. Molecular genetic techniques hold the promise of isolation and characterization of the genetic defect(s) in familial AD (FAD). Recently, chromosome 21 has been implicated as the potential site of an autosomal dominant defect in some but not necessarily all FAD pedigrees. However, the results of recent genetic epidemiologic studies suggest that progress in the molecular genetic approach to AD will be difficult.