Alterations in cellular immune function are associated with depression and have been related to changes in endocrine function. We carried out a study to: (i) reliably assess the hypothalamic-pituitary-adrenal (HPA) axis function in treatment resistant depression (TRP); (ii) evaluate whether depression was associated with changes on T-cell proliferation and cytokine production; and (iii) assessed the sensitivity of lymphocytes to glucocorticoids (GC)s in vitro. Thirty-six pharmacologically treated inpatients diagnosed with TRP and 31 healthy controls took part in the study. Salivary cortisol was measured hourly from 0800 to 2200 h both before and after dexamethasone (DEX) intake and the patients were classified into HPA axis suppressors and nonsuppressors. The following were measured in vitro: (a) phytohemagglutinin-induced T-cell proliferation; (b) cytokine production (interleukin-2 and tumor necrosis factor-alpha, TNF-alpha); and (c) lymphocyte sensitivity to both cortisol and DEX. Basal morning cortisol levels from patients and controls did not differ nor did their T-cell proliferation and cytokine production. Ten out of 36 patients were classified as nonsuppressors and presented a significantly higher post-DEX salivary cortisol levels than suppressors, 82.0 vs 8.9 nM/l/h (p <0.001). Cells of nonsuppressors produced significantly less TNF-alpha compared to suppressors, 299.8 vs 516.9 pg/ml (p < 0.05). Remarkably, GC-induced suppression of lymphocyte proliferation and cytokine production were generally less marked in depressives compared with controls. Our data indicate that alterations in immune function and steroid regulation associated with depression are not related to elevated basal levels of cortisol and suggest that lymphocyte steroid resistance may be associated with TRP.