The 12-month outcome of 103 elderly depressed patients treated by psychiatrists in Perth, Western Australia, was good in 32-47% of cases, depending upon the stringency of the outcome criteria used. A higher than expected mortality was found, especially in men. No clear association was found between 12-month outcome and a variety of clinical variables at index admission, including chronic physical illness, severity of depression, and severe life events during follow-up. As the statistical power of this and other studies is small, it would be unwise at present to draw definite conclusions about the role of possible risk factors in the prognosis of depressed elderly. Vigorous treatment of the depressive illness and adequate after-care are important.