A 12-month prospective study of a randomly selected group of acute geriatric medical inpatients was carried out. Twenty-three of the original sample of 72 were diagnosed as cases of depression on entry into the study and commenced on treatment with fluoxetine 20 mg mane. Follow-up at 3 months showed a significantly increased death rate in the depressed group (p < 0.05) which was reversed in those surviving to 12 months, this trend to a lower mortality not reaching significance. Two patients did not continue their antidepressants after leaving hospital and remained depressed at 3-month follow-up. The effect of treatment versus no treatment was significant (p < 0.03). The incidence of depression in the psychologically well group was 14% over the 12-month period, understandably high when seen in the context of their continuing physical problems. None of the treated depressed group relapsed during the study period.