Objective To report mortality risks of dementia based on national hospital registry data, and to put these risks into perspective by comparing them with those in the general population and following cardiovascular diseases. Design Prospective cohort study from 1 January 2000 through 31 December 2010. Setting Hospital-based cohort. Participants A nationwide hospital-based cohort of 59 201 patients with clinical diagnosis of dementia (admitted to a hospital or visiting a day clinic) was constructed (38.7% men, 81.4 years (SD 7.0)). Main outcomes and measures 1-year and 5-year age-specific and sex-specific mortality risks were reported for patients with dementia visiting a day clinic compared with the general population; for patients hospitalised with dementia compared with patients hospitalised for acute myocardial infarction (AMI), heart failure or stroke, these were presented as absolute and relative risks (RRs). Results 1-year mortality was 38.3% in men and 30.5% in women. 5-year risk was 65.4% and 58.5%, respectively. Mortality risks were significantly higher in patients with dementia admitted to the hospital than in those visiting a day clinic (1-year RR 3.29, 95% CI 3.16 to 3.42; and 5-year RR 1.79, 95% CI 1.76 to 1.83). Compared with the general population, mortality risks were significantly higher among patients visiting a day clinic (1-year RR for women 2.99, 95% CI 2.84 to 3.14; and for men 3.94, 95% CI 3.74 to 4.16). 5-year RRs were somewhat lower, but still significant. Results were more pronounced at younger ages. Mortality risks among admitted patients were comparable or even exceeded those of cardiovascular diseases (1-year RR for women with dementia vs AMI 1.24, 95% CI 1.19 to 1.29; vs heart failure 1.05, 95% CI 1.02 to 1.08; vs stroke 1.07, 95% CI 1.04 to 1.10). 5-year RRs were comparable. For men, RRs were slightly higher. Conclusions Dementia has a poor prognosis as compared with other diseases and the general population. The risks among admitted patients even exceeded those following ...