Background. Little is known about the prognosis of depression in older patients in general practice or the community. Objectives. To summarize available evidence on the course and prognostic factors of depression in older persons. Methods. We conducted a systematic, computerized search of Medline and PsycINFO. Manual search of references of included studies were done. Studies potentially eligible for inclusion were discussed by two reviewers. Methodological quality was independently assessed by two reviewers. Data regarding selection criteria, duration of follow-up, outcome of depression and prognostic factors were extracted. Results. We identified 40 studies reporting on four cohorts in general practice and 17 in the community. Of all, 67% were of high quality. Follow-up was up to 1 year in general practice and up to 10 years in the community. Information on treatment was hardly provided. About one in three patients developed a chronic course. Five cohorts used more than two measurements during follow-up, illustrating a fluctuating course of depression. Using a best evidence synthesis we summarized the value of prognostic indicators. General practice studies did not provide strong evidence for any factor. Community studies provided strong evidence for an association of baseline depression level, older age, external locus of control, somatic co-morbidity and functional limitations with persistent depression. Conclusion. Within the older population, age seems to be a negative prognostic factor, while older people are more likely to be exposed to most of the other prognostic factors identified.