In this prospective multicentre trial, 90 patients undergoing autologous stem cell transplantation (ASCT) were randomised to receive (n=43) or not receive (n=47) amifostine 910 mg/m(2) prior to melphalan 200 mg/m(2). Patients were monitored for regimen-related toxicity, engraftment, supportive care, response and survival. Both groups underwent ASCT at a median of 8 months from diagnosis and were matched for disease characteristics, prior therapy and pre-ASCT disease responsiveness. Amifostine infusional side-effects were frequent, occurring in 65% of patients, but of mild severity. Amifostine use was associated with a reduction in the median grade of oral mucositis (1 vs 2, P=0.01) and the frequency of severe (WHO grades 3 or 4) mucositis (12 vs 33%, P=0.02), but no reduction in the requirement for parenteral nutrition or analgesic use. Conversion to complete remission post-ASCT occurred in 30 and 14% of the amifostine and control groups, respectively (P=0.09). With a median follow-up of 35 months, there was no statistically significant difference between the median progression-free or overall survival times for the two groups. We conclude that amifostine can be safely administered prior to high-dose melphalan and significantly reduces the frequency and severity of therapy-induced oral mucositis.