To determine the usefulness of the DST in differentiating depression from dementia, the test was administered to three diagnostic groups of psychogeriatric patients: depression; dementia; and dementia with depression. Clinical assessments were supplemented by ratings on the HRSD and SCAG, as well as by EEG and CT. All three groups showed a high incidence of abnormal DST results unrelated to presence or severity of affective symptoms, but showing a better association with SCAG and its ‘organic’ subsets. The mechanism(s) underlying these abnormal results may reflect organic brain disease. The usefulness of the DST in differentiating depression from dementia in the elderly was not confirmed.