Abstract. Intravenous iron (iv.Fe) is used to optimize response to recombinant human erythropoietin (r-HuEPO) in ESRD, but no consensus exists with respect to the best regimen to avoid transferrin "oversaturation," oxidative stress, and the occurrence of non-transferrin-bound iron (NTBI). Iv.Fe was stopped for 1 wk in 35 hemodialysis (HD) patients who were routinely receiving iv.Fe and r-HuEPO. The iv.Fe group received 100 mg of ferric saccharate (Venofer) at the end of the first HD session, whereas the time-control group was treated under the same conditions but received no iv.Fe. Serum samples were taken before the first HD session, immediately and 60 min after iv.Fe administration, and before the next HD session. Sera were analyzed for NTBI and peroxides; transferrin saturation was analyzed by urea-PAGE and Western blot. In an in vitro model system with HepG2 cells, the effects of ESRD serum on the labile iron pool (LIP) were assayed using the fluorescence calcein assay. NTBI significantly increased after iv.Fe-administration and returned to baseline values before the next HDsession. There was a shift from apo-to monoferric transferrin, but no "oversaturation" of transferrin after iv.Fe-treatment. Peroxides increased in both groups after HD. Hemodialysis decreased bioavailable iron for the LIP in HepG2-cells, whereas serum of iv.Fe-treated HD patients highly increased the LIP in these cells. A total of 100 mg of iv.Fe led to NTBI generation but not to an oversaturation of transferrin. Peroxide concentrations significantly increased during HD but were not correlated to iv.Fe administration and seemed to result from other sources of oxidative stress related to HD. NTBI can enter liver cells and increase the potentially harmful LIP.