Purpose: To assess and if needed improve the compliance of participating institutions to the radiotherapy guidelines of the EORTC AMAROS trial 10981/22023 comparing axillary radiotherapy to axillary surgery in sentinel node positive patients with early stage breast cancer. Materials and methods: A transverse contour and a frontal view radiograph of the axillary region of a 'dummy' patient were sent to all institutions intending to participate in the trial with the request to produce a radiotherapy treatment plan according to the protocol guidelines. Additional information on dose prescription, the treatment technique and field matching with breast fields and internal mammary lymph node fields was requested in a questionnaire. Results: Eighteen institutions have performed the dummy run. At first assessment, the dose was not specified according to the protocol in seven cases, while two institutions did not comply with the dose prescription of 50 Gy in 25 fractions. Dose heterogeneity was over 20% in 10 institutions, caused by the use of a two-field technique in eight cases. Ten institutions did not apply special techniques to obtain non-overlapping match planes. In 10 cases, one or more field borders or blocks were positioned incorrectly. Following recommendations from the quality assurance committee given to the participating institutions on an individual basis, 10 institutions adapted their technique. Thereafter, 16 institutions could be accepted for trial participation. Conclusions: A number of potential protocol deviations were found at first assessment. Since recommendations led to a large number of adaptations by the participants, a considerable improvement in protocol compliance and inter-institutional consistency was achieved. © 2003 Elsevier Ireland Ltd. All rights reserved.