Post-mortem analyses suggest that atherosclerosis more often contributes to late-onset dementia than hitherto expected. We set out to further unravel the relation between atherosclerosis and cognitive impairment. We therefore tested the hypothesis that the number of cardiovascular pathologies is positively associated with cognitive impairment in elderly subjects, and that the smaller number of cardiovascular pathologies in women explains the better cognitive function of elderly women. Within the Leiden 85-plus Study, we assessed the atherosclerotic burden by counting the number of cardiovascular pathologies in the medical histories of a population-based sample of 599 subjects aged 85 years (response 87%). Significantly more men than women had a history of cardiovascular pathologies (67% compared to 59%, P<0.001). In addition, cognitive function was assessed. All subjects completed the Mini-Mental State Examination (MMSE). Cognitive speed and memory were determined with specific neuro-psychological tests in those with a MMSE-score above 18 points. There was a highly significant dose-response relationship between the number of cardiovascular pathologies and cognitive impairment for both men and women. The median MMSE-score was 26 points in subjects without cardiovascular disease and decreased to 25 points for subjects who had two or more cardiovascular pathologies (P for trend =0.003). Similar associations were found for cognitive speed but not for memory. Our data confirm that in old age atherosclerosis significantly contributes to cognitive impairment. Since treatments for atherosclerosis appear to be particularly effective in elderly people, we consider our finding of utmost clinical importance in possibly preventing cognitive impairment and late-onset dementia.