Abstract The dexamethasone suppression test (DST), the thyrotropin releasing hormone (TRH) test and the Newcastle II depression rating (Nil) were compared with the clinical diagnosis and evaluated in 61 patients fulfilling the criteria of an affective disorder according to the DSM-III classification. A statistically significant correlation between clinical diagnosis and DST as well as Nil, but not between clinical diagnosis and TRH test, was found. There was do correlation between DST and the severity of depression according to the Hamilton depression rating. The nosographic and the diagnostic specificities and sensitivities for the DST, TRH test and Nil and combinations hereof were correlated with the clinical diagnosis. Combining DST and Nil, a nosographic sensitivity of 50 % and a nosographic specificity of 84 % were found. Correspondingly, the diagnostic sensitivity was 43 % and the diagnostic specificity was 88 %. The DST and the TRH test were found of no value in the prediction of the response to antidepressive treatment. Mainly because of a low diagnostic sensitivity the Nil, the DST and the TRH test are of limited value in the diagnosis of depressive disorders.