This prospective, randomised pilot study was designed to evaluate safety, feasibility and efficacy of recombinant human erythropoietin (rhEPO) in the prevention and treatment of chemotherapy-induced anaemia in children with solid tumours. 20 children (age 4-18 years) undergoing cyclic combination chemotherapy were randomised either to a control group or to receive rhEPO at a dose of 150 U/kg/dose subcutaneously three times/week for a minimum of 12 weeks or three chemotherapy cycles. Of 15 evaluable patients, 8 were randomised to the rhEPO group and 7 to the control group. RhEPO-treated patients showed an increase in the haematocrit over the first 8 weeks of therapy, with a significantly higher mean haematocrit at week 8 (33.2 +/- 2.1% versus 39.3 +/- 4.2% in the control and rhEPO groups, respectively, P < 0.05). Similarly, significantly higher haemoglobin concentrations could be demonstrated in the rhEPO group by week 8 (11.06 +/- 1.35 g/dl versus 13.11 +/- 1.13 g/dl in the control and rhEPO groups, respectively, P < 0.05), with higher precycle haemoglobin before chemotherapy cycles 3 and 4 and higher midcycle haemoglobin between cycles 3 and 4. There was a trend towards a reduction of transfusion requirements during the 3rd month of therapy in rhEPO patients. The results of this pilot study indicate a significant benefit of rhEPO in children treated with intensive combination chemotherapy regimens. Further studies should target issues such as appropriate dosing, timing and duration of rhEPO therapy in children with cancer.