Background: The use of benzodiazepines in relation to cognitive decline remains an area of controversy in aging populations. Objective: This study aims to evaluate the risk of cognitive impairment–not dementia (CIND), Alzheimer disease (AD), and all-cause dementia with benzodiazepine use. The effect modification by sex was also investigated. Methods: Data come from the Canadian Study of Health and Aging, a 10-year multicentric study involving 10 263 participants randomly selected, 65 years and older, living in the community and in institutions. Current exposure to benzodiazepines was assessed in a face-to-face interview or self-reported in a questionnaire. Cox proportional hazard regression models, using age as time scale, were conducted to estimate hazard ratios, with adjustment for sex, education, smoking, alcohol intake, depression, physical activity, nonsteroidal anti-inflammatory drug use, and vascular comorbidities. Results: Data sets included 5281 participants for dementia as the outcome, 5015 for AD, and 4187 for CIND. Compared with nonusers, current use of benzodiazepines was associated with an increased risk of CIND (hazard ratio = 1.36; 95% CI = 1.08-1.72) in the simplest model. Results remained similar in the fully adjusted model (hazard ratio = 1.32; 95% CI = 1.04-1.68). There was no association between benzodiazepine use and the risk of dementia or AD. All these effects were similar between men and women. Conclusion and Relevance: Benzodiazepine use in older people from the general population is related to subsequent occurrence of cognitive dysfunction but not implicated in the pathogenesis of dementia or AD. Caution should be exercised when prescribing benzodiazepines to preserve global cognitive function.