A re-analysis of the data from 11 case-control studies was performed to investigate the association between head trauma and Alzheimer's disease (AD). To increase comparability of studies, exposures were limited to head trauma with loss of consciousness (hereafter referred to as 'head trauma') and comparisons were restricted to community (versus hospital) controls. Test for heterogeneity across studies was negative; consequently, data were pooled in subsequent analyses. The pooled relative risk for head trauma was 1.82 (95% confidence interval: 1.26-2.67). Stratified analyses showed stronger associations in cases without a positive family history of dementia and in males (versus females). Adjustment of the pooled relative risk for family history of dementia, education and alcohol consumption did not alter significantly the association between head trauma and AD. There was no interaction effect between head trauma and family history of dementia, suggesting that these risk factors operate independently. Mean age of onset was not significantly different in cases with a history of head trauma compared to cases without such a history. The findings of the pooled analysis support an association between reported head trauma and AD.