In unilateral total knee replacement (TKR), perioperative blood loss, low transfusion thresholds and short hospital stay result in patients being discharged with low haemoglobin (Hb). We assessed the effects of perioperative administration of intravenous iron, with or without erythropoietin, plus a restrictive transfusion threshold (Hb < 80 g L(-1)) both on transfusion rate and recovery from post-operative anaemia. TRK patients received iron sucrose (2 x 200 mg per 48 h, iv) (Group IVI, n = 129). Patients with admission Hb < 130 g L(-1), also received erythropoietin (1 x 40 000 IU, sc) (Group EPO, n = 19). Perioperative clinical and laboratory data were obtained. Mean Hb loss was 36 g L(-1), but only seven patients were transfused (5%). Pre-operatively, 66 (45%) patients did not have enough stored iron to compensate Hb loss. At post-operative day 30, only 15% were anaemic, 70% of Hb loss and 92% of pre-operative Hb were recovered and ferritin increased by 73 microg L(-1) (P < 0.01), although erythropoietic response was higher in patients receiving erythropoietin (P < 0.05). No adverse effects of iron sucrose or erythropoietin were witnessed. This protocol seems to reduce allogeneic blood transfusion rate and may hasten the recovery from post-operative anaemia in TKR patients, without depleting iron stores. Further studies are needed to ascertain which patients may benefit of extended intravenous iron and/or erythropoietin administration.