Despite the fact that people age 65 and older have the highest rates of suicide of any age-group, late-life suicide has a low prevalence, making it difficult to conduct prospective studies. The authors examined risk factors for late-life suicide on the basis of general information collected directly from older subjects participating in a community-based prospective study of aging, the Established Populations for Epidemiologic Studies of the Elderly. Demographic variables, presence of a relative or friend to confide in, alcohol use, and sleep quality were assessed at baseline interview. Baseline and follow-up data were used to determine physical, cognitive, and affective functioning, as well as medical status. Of 14,456 people, 21 committed suicide over the 10-year observation period. Depressive symptoms, perceived health status, sleep quality, and absence of a relative or friend to confide in predicted late-life suicide. Suicide victims did not have greater alcohol use and did not report more medical illness or physical impairment. This study provided additional information about the context of late-life depression that also contributes to suicidal behavior: poor perceived health, poor sleep quality, and limited presence of a relative or friend to confide in.