Late-life depression is highly heterogeneous in clinical presentation, and is also commonly resistant to treatment. While some cases are a continuation of the chronic course of illness beginning in early adulthood, a large number of persons will have a first episode of depression in later life following alife-time of relatively good mental health. While incident cases of major depression tend to decrease with age, the number of persons with clinically significant depressive symptomatology rises. À distinction has often been made between early-onset and late-onset depression, however, there is no conclusive evidence to suggest these are distinct clinical entities. On the other hand observations from a fifteen year prospective population study of psychiatric disorder in the elderly (the ESPRIT Study) supports the alternative idea that depression may be divided into sub-types according to postulated aetiology; for example depression with a strong genetic component, related to hormonal changes, the consequence of trauma; the result of cerebrovascular insult. Exposure to these putative causes may be more common at different points in the life span, thus suggesting age-differences. Our research further suggests that even cases of depression appearing for the first time in late-life, may be initially triggered by risk factors occurring decades before. Our findings suggest, for example, that childhood events may lead to changes in the biology of stress management, which continue throughout life, increasing vulnerability to depression and persisting even after effective treatment of symptoms. Together these observations suggest it may be more meaningful to classify depression in the elderly according to probable principle precipitating factors rather than age.