Objective: To determine if older men who use computers have lower risk of developing dementia.Methods: Cohort study of 5506 community-dwelling men aged 69 to 87 years followed for up to 8.5 years. Use of computers measured as daily, weekly, less than weekly and never. Participants also reported their use of email, internet, word processors, games or other computer activities. The primary outcome was the incidence of ICD-10 diagnosis of dementia as recorded by the Western Australian Data Linkage System.Results: 1857/5506 (33.7%) men reported using computers and 347 (6.3%) received a diagnosis of dementia during an average follow up of 6.0 years (range: 6 months to 8.5 years). The hazard ratio (HR) of dementia was lower among computer users than non-users (HR = 0.62, 95%CI = 0.47–0.81, after adjustment for age, educational attainment, size of social network, and presence of depression or of significant clinical morbidity). The HR of dementia appeared to decrease with increasing frequency of computer use: 0.68 (95%CI = 0.41–1.13), 0.61 (95%CI = 0.39–0.94) and 0.59 (95%CI = 0.40–0.87) for less than weekly, at least weekly and daily. The HR of dementia was 0.66 (95%CI = 0.50–0.86) after the analysis was further adjusted for baseline cognitive function, as measured by the Mini-Mental State Examination.Conclusion: Older men who use computers have lower risk of receiving a diagnosis of dementia up to 8.5 years later. Randomised trials are required to determine if the observed associations are causal.