Objectives This study aims to determine the prevalence of potentially undetected dementia among institutional care facility residents in Ontario, Canada, and to identify factors associated with undetection. Methods We utilized a population‐based secondary data analysis approach, pertaining to data from the Canadian Institute for Health Information's Continuing Care Reporting System, 2009–2011. Potentially undetected dementia was defined as having severely impaired cognitive function and requiring extensive assistance on activity of daily living (ADL) but no records of dementia diagnoses. Cognitive function was measured by the Cognitive Performance Scale (CPS), 0 (intact) to 6 (very severe impairment), and ADL by a hierarchy scale, 0 (independent) to 6 (total dependence). Results Of the 242,957 residents who had no records of dementia diagnoses, 11.6% (n = 28,078) had a CPS score ≥4 (severe impairment or higher) and ADL score ≥3 (required extensive assistance or more). Data from 11,614 demented residents with corresponding CPS and ADL scores were used for comparison. Residents without dementia diagnosis were younger (77 vs. 84 years), more likely to have never married (20% vs. 6%), and have longer admission (4 vs. 2.8 years). The most significant factors for no diagnoses were never married (adjusted odds ratio = 2.1, 95% confidence interval [CI] = 1.91–2.29), admitted to hospital‐based facilities (adjusted odds ratio = 1.58, 95% CI = 1.48–1.69), presence of schizophrenia (adjusted odds ratio = 1.43, 95% CI = 1.22–1.69), depression (adjusted odds ratio = 1.23, 95% CI = 1.16–1.29), and diabetes mellitus (adjusted odds ratio = 1.32, 95% CI = 1.26–1.40). Conclusions A large number of residents who had poor cognitive function and inadequate ADL ability did not have dementia diagnoses on record. Social and comorbid conditions were contributing factors to potentially undetected dementia. Copyright © 2013 John Wiley & Sons, Ltd.