Lithium augmentation is a well established strategy for treatment-resistant depression. The exact mode of its action is unknown, but an enhancement of serotonergic transmission is hypothesized. The authors investigated changes in the hypothalamic-pituitary-adrenocortical (HPA) system during lithium augmentation and their correlation to clinical response by means of the combined dexamethasone/CRH test (DEX/CRH test). Thirty patients with unipolar major depressive episodes (DSM IV) who had not responded to an adequate trial with an antidepressant were assessed on the day before lithium augmentation (baseline) with the DEX/CRH test (pretreatment with 1.5 mg dexamethasone p.o. at 11 P.M. and CRH stimulation at 3 P.M. on the next day). Twenty-four patients were re-assessed after response was determined or, in cases of non-response, four weeks after initiation of lithium augmentation. Response to lithium augmentation was measured by weekly ratings on the Hamilton Depression Rating Scale (HDRS 17-item version). Response was defined as a DeltaHDRS of > or =50% and an endpoint score of < 10. Patients had a significantly higher ACTH and cortisol response to CRH stimulation during lithium augmentation compared with the values at baseline. There was no difference in ACTH and cortisol reaction between responders and non-responders to lithium augmentation. This increase is in contrast to the known normalization of HPA-axis overdrive after treatment with a tricyclic antidepressant like amitriptyline. Because the effect was independent of response status we suggest that this increase reflects an effect of lithium that is independent from the psychopathological state or its change. This effect might be explained by the serotonergic effects of lithium.