Linkage studies have implicated a broad region on chromosome 10q in Alzheimer's disease (AD). A recent genetic association study has provided evidence that polymorphism in the gene encoding alpha-3 catenin (CTNNA3, referred to previously as VR22 and also known as alpha-T catenin) may underlie linkage signals. Here, to investigate this finding, markers that previously exhibited maximum evidence of association have been tested in Swedish and Scottish AD case-control samples. Across models of disease risk and in relation to multiple quantitative indices of AD pathology (CSF A beta 42 and tau levels, age-at-onset, MMSE scores, and measures of senile plaque density) no evidence was found supporting a role for these particular variants in AD. More detailed studies of regional linkage disequilibrium structure around CTNNA3 will likely be required to determine whether sequence variation in this region impacts AD.