The aim of this study was to assess any inconsistency with the protocol guidelines for preoperative radiotherapy for rectal cancer among radiotherapy institutions participating in the framework of a multicentre phase-III European Organization for Research and Treatment of Cancer (EORTC) clinical trial. Twelve radiotherapy departments with more than 10% of the evaluable patients recruited in the trial, were invited to participate in this individual case review. Participating institutions were asked to send five full patient records with the chemotherapy charts, surgical and pathology reports, radiation treatment charts, treatment planning calculations, computed tomography (CT) scans, portal images and follow-up charts. The sample of the 5 patients per institution was randomly selected at the EORTC Data Center. All 12 departments participated. In 21 (35%) of the cases, a three-field technique with two lateral opposed wedge fields and a posterior field was used, while in 39 (65%) of the cases a four-field pelvic box technique was used. All participants used linear accelerators with a minimum of 6 MeV photon-beam energy. In general, the patients were eligible, documentation of clinical data was fair to good and there were no systematic major protocol deviations. The actual total dose was 44.7±a standard deviation (S.D.) of 4.6 Gy. Some variation was found in the fraction size. All institutions complied with the protocol in specifying the reference dose at the ICRU point. The clinical target volume (CTV) drawn on the CT scan was narrow in 7 (12%) cases, but eventually the actually treated volumes in terms of planned treatment volume (PTV) were correct. In two institutions, although the CTV was drawn correctly, the fields appeared to be narrow especially in cranio-caudal direction. Variations in treated volumes and total radiation dose were encountered in the individual case review. By providing recommendations early during the course of the trial, we expect to improve the inter-institutional consistency and to promote a high quality treatment in all of the participating institutions. © 2002 Elsevier Science Ltd. All rights reserved.