Studies using diverse methods have documented frontostriatal and limbic dysfunction occurring in late-life depression. Although such impairments may result from aging-induced brain changes unrelated to depression, there are at least two reasons to suggest that they play a pathogenetic role in geriatric depression. First, frontostriatal dysfunction has been identified in at least some younger depressed subjects without known neurological abnormalities. Second, frontostriatal dysfunction may be associated with poor short- and long-term outcomes of late-life depression. Relating frontostriatal and limbic dysfunction to the course of late-life depression is an appropriate way for investigating its pathophysiological relevance, given that no biological test can be used as a validating criterion. However, this approach has experimental limitations. Studies of the course of late-life depression may be influenced by selective survival of depressed patients with favorable prognosis; factors peripherally related to the biology of depression, for example, physical handicaps; and clinical factors with unclear relationship to specific biological abnormalities, for example, personality disorders. Nonetheless, studies comparing depressed patients with control subjects complemented with studies of course of illness can bring to bear the rapidly evolving cognitive-neuroscience and brain-imaging techniques in an investigation of the networks responsible for predisposing, precipitating, and perpetuating late-life depression.