To determine whether recombinant human erythropoietin improves the quality of life and exercise capacity of anaemic patients receiving haemodialysis. A double blind, randomised, placebo controlled study. Eight Canadian university haemodialysis centres. 118 Patients receiving haemodialysis aged 18-75 with haemoglobin concentrations less than 90 g/l, no causes of anaemia other than erythropoietin deficiency, and no other serious diseases. Patients were randomised to three groups to receive placebo (n = 40), erythropoietin to achieve a haemoglobin concentration of 95-110 g/l (n = 40), or erythropoietin to achieve a haemoglobin concentration of 115-130 g/l (n = 38). Erythropoietin was given intravenously thrice weekly, initially at 100 units/kg/dose. The dose was subsequently adjusted to achieve the target haemoglobin concentration. All patients with a serum ferritin concentration less than 250 micrograms/l received oral or intravenous iron for one month before the study and as necessary throughout the trial. Scores obtained with kidney disease questionnaire, sickness impact profile, and time trade off technique; and results of six minute walk test and modified Naughton stress test. The mean (SD) haemoglobin concentration at six months was 74 (12) g/l in patients given placebo, 102 (10) g/l in those in the low erythropoietin group, and 117 (17) g/l in those in the high erythropoietin group. Compared with the placebo group, patients treated with erythropoietin had a significant improvement in their scores for fatigue, physical symptoms, relationships, and depression on the kidney disease questionnaire and in the global and physical scores on the sickness impact profile. The distance walked in the stress test increased in the group treated with erythropoietin, but there was no improvement in the six minute walk test, psychosocial scores on the sickness impact profile, or time trade off scores. There was no significant difference in the improvement in quality of life or exercise capacity between the two groups taking erythropoietin. Patients taking erythropoietin had a significantly increased diastolic blood pressure despite an increase in either the dose or number of antihypertensive drugs used. Eleven of 78 patients treated with erythropoietin had their sites of access clotted compared with only one of 40 patients given placebo. Patients receiving erythropoietin were appreciably less fatigued, complained of less severe physical symptoms, and had moderate improvements in exercise tolerance and depression compared with patients not receiving erythropoietin. At the doses used in this trial there was a higher incidence of hypertension and clotting of the vascular access in patients treated with erythropoietin.