In the AML 8A study patients were treated with remission-induction therapy followed by one consolidation course. Patients in complete remission (CR) were randomised between autologous bone marrow transplantation (ABMT) and a second intensive consolidation course, except for those with a histocompatible sibling donor, who received allogeneic bone marrow transplantation (alloBMT). This analysis was performed to determine whether centres which only performed induction and consolidation therapy, achieved similar results as centres who also performed transplantation. 542/676 (80%) from transplantation centres and 150/194 (77%) from referring centres achieved CR, with an early death rate of 5% and 11%, respectively (P = 0.01). 66% of patients with a donor from transplantation centres received alloBMT in first CR compared with 57% from referring centres (P = 0.2). Transplantation centres randomised 64% of patients without a donor, referring centres 47% (P = 0.04). The full protocol treatment was completed by 275/542 (51%) and 61/150 (41%) patients, respectively (P = 0.04). The overall survival rate at 6 years from diagnosis was 34% and 36%, respectively (P = 0.9). In conclusion, the type of centre did not appear to have an influence on overall survival. The feasibility of the study was acceptable for both types of centres. The referring centres applied more selection for transplantation. Despite a more intensive second-line treatment at transplantation centres, the overall outcome remained similar to that of referring centres.