Patients receiving epoetin therapy show wide variability in their responsiveness to the drug. Many factors may be responsible for this, particularly iron deficiency, acute infection and under-dialysis. Even after excluding factors known to cause resistance to epoetin, the marked variability in sensitivity to epoetin remains. The exact mechanism of this effect is unclear. It is, however, recognized that uraemia is a chronic inflammatory state, with some patients showing quite significantly increased laboratory markers of inflammation and immune activation. It is also known that chronic inflammation can modify the process of erythropoiesis, and this is probably mediated via pro-inflammatory cytokines such as interleukin-1 (IL-1), tumour necrosis factor-alpha (TNF-alpha) and interferon-gamma (IFN-gamma). It is hypothesized, therefore, that some patients showing resistance to epoetin may have enhanced levels of immune activation, causing increased release of pro-inflammatory cytokines in the bone marrow. This has been investigated by studying T-cell phenotypes by flow cytometry, along with cytokine release from T cells and monocytes in 'good' and 'poor' responders to epoetin. Poor responders were found to have significantly reduced CD28 expression on both CD4(+) and CD8(+) cells, enhanced IL-10 generation from peripheral blood mononuclear cells (PBMCs), higher plasma IL-12 levels and enhanced TNF-alpha release from PBMCs. The patients in this study, who were followed-up for the subsequent 24 months, had a considerably lower survival if they were poor responders (54% vs 88% for good responders; P<0.05). Further work in this area is required to confirm or contest the hypothesis that epoetin resistance is due to enhanced levels of immune activation.