BACKGROUND: Several risk factors of depression have been identified in retrospective as well as some prospective studies in the elderly. Confirmation in independent samples is needed. The present follow-up study prospectively investigated risk factors of depression in an elderly German sample. METHODS: One thousand four hundred and thirty-one subjects from a family study were re-investigated after 4.7 +/- 2.5 years. Bivariate and multivariate forward logistic regression analyses were used to identify risk factors of the development of new depression in the elderly. RESULTS: Risk factors of a new depressive episode in 1408 elderly without current depression were age, female gender, a previous depression, subjective memory impairment, previous anxiety and somatoform disorders. The presence of dementia or mild cognitive impairment were significant risk factors in bivariate, but not multivariate analysis controlling for possible confounding. Risk factors of a first geriatric depressive episode were age, gender and subjective memory impairment; age remained the only significant risk factor in multivariate analysis. CONCLUSIONS: This investigation confirms previous studies from other countries concerning the relevance of risk factors for depression in the elderly. The knowledge of risk factors might help identify subjects at increased risk of depression for early intervention approaches. Elderly with a history of previous depression carry the highest risk.