The dexamethasone suppression test (DST) was administered shortly after admission to 102 consecutive in-patients with a Hamilton depression score ≥16. Post-dexamethasone cortisol exceeded 6pg/dl in 16 cases, and levels correlated significantly with Hamilton scores; with the AMP syndromes ‘hypochondria’, ‘apathy’ and ‘catatonia’; and with the IMPS ‘retarded depressive’ syndrome. The criterion of suppression/non-suppression did not distinguish significantly between diagnostic categories (RDC or ICD), nor between endogenous and neurotic depression. (Newcastle scale). Both base-line and post-dexamethasone cortisol levels were reduced by prior treatment with minor tranquillisers, but not by major tranquillisers or antidepressants. DST results cannot be used as straightforward indicators of prognosis.