The prognostic significance of urinary incontinence (UI) was evaluated using the Helsinki (Finland) Aging Study Database, in which a random sample of persons of three age cohorts (75, 80 and 85 years, n=649) was followed for 5 years. The prevalence rates of UI in the three age groups were 10.5%, 20.4% and 28.6%, respectively. UI was also frequent in healthy aged subjects (15.6%), but most common in hospitalized patients (35.3%), in those with severe depression (51.8%), previous stroke (37.6%) or dementia (48.8%). UI had age- and gender-adjusted risk ratios (95% C.I.) for mortality of 1.75 (1.33-2.30) and for long-term institutional care of 2.37 (1.30-3.84). However, after controlling for the presence of dementia, the risk ratios were no longer significant. Thus, UI predicts both death and long-term institutionalization in the general aged population. However, this observation is mainly explained by the close association of UI with dementia.