A blunted thyrotropin (TSH) response is a predictor of a good response to antidepressant drug treatment in depressives and neuroleptic treatment in paraphrenic patients (Larger et al 1986). The aim of the following study was to elucidate possible relationships between different endocrine systems and to shed light on the pathogenetic hypotheses of TSH-blunting. In order to evaluate especially hypothalamic activity in severe depression we were interested in the vasopressin system as another hormonal system underlying hypothalamic control. Thirty-four patients who met the criteria for major depression according to DSM-III-R were subjected to the thyrotropin-releasing hormone (TRH) test. We also took baseline readings of the cortisol, neurophysinI (hNpI, reflecting vasopressin plasma levels), and neurophysinII (hNpII, reflecting oxytocin plasma levels) levels. Likelihood ratio tests were done with logistic regression models to analyze the phenomenon of TSH-blunting. We observed that the likelihood of a blunted TSH response increases with higher levels of hNpI and low levels of cortisol, but is unrelated to hNpII levels.