The development and course of depression is causally linked to impairment of central regulation of the hypothalamic-pituitary-adrenocortical (HPA) system. Previous research documented that the combined dexamethasone/corticotropin-releasing hormone (DEX/CRH) test identifies HPA dysfunction with high sensitivity. We evaluated the predictive validity for medium-term outcome of the cortisol response in the combined DEX/CRH test in 74 remitted patients previously suffering from major depressive disorder. Of the 74 patients, 61 remained in stable remission and 13 relapsed during the first 6 months after discharge from the hospital. Although the cortisol and ACTH responses in the DEX/CRH test did not differ between the two groups of patients on admission, the responses differed significantly just before discharge (P< 0.05). We defined two dichotomous variables as prediction rules indicating (1) the change between admission and discharge in the cortisol response to the DEX/CRH test, and (2) the effect of the CRH infusion on cortisol as compared to the baseline level in the DEX/CRH test prior to discharge only. An elevated cortisol response in the DEX/CRH test was correlated with a four- to six-fold higher risk for relapse than in individuals with a normal cortisol response. The two proposed rules for predicting relapse within the first 6 months after discharge could be optimized by including age and gender. Hence, an exaggerated cortisol response in the combined DEX/CRH test predicts the recurrence of depressive psychopathology. The test performance can be further optimized if gender and age are taken into account.