Genetic association has recently been reported between alleles in the alpha-2-macroglobulin (A2M) gene and the occurrence of Alzheimer's disease (AD) in familial and sporadic samples. We have investigated the A2M intronic deletion polymorphism in a case-control study of 295 unrelated clinic and community-based AD cases, and compared these to a sample of 113 unrelated control individuals recruited as part of an epidemiological study. Our results show no association between A2M and AD in either case sample. Furthermore, A2M is not predictive of AD in an interactive fashion when considering APOE, race or gender. In a subset of our larger sample we have also investigated the A2M Val1000lle polymorphism, and again find no evidence for association. We conclude that there is no genetic association between A2M and AD in our case-control sample.