Histopathological audit of positive circumferential resection margins (CRMs) can be used as a surrogate measure of the success of rectal cancer treatment. We audited CRM involvement in rectal cancer patients and the impact of the multidisciplinary team (MDT) on implementing a magnetic resonance imaging (MRI)-based preoperative treatment strategy. Data were collected on all newly diagnosed rectal cancer patients treated in our network between January 1999 and December 2002. Data were analysed for MRI prediction and histopathological assessment of CRM together with the MDT meeting treatment decisions. The CRM+ve rate of those discussed at MDT vs those not discussed were compared. We re-audited the CRM+ve rates 1 year after introducing a policy of mandatory preoperative MRI-based MDT discussion. Of the 298 patients diagnosed with rectal cancer, 39 (13%) were deemed palliative, 178 underwent surgery alone and 81 underwent neoadjuvant therapy. Of these, 62 out of 178 patients underwent surgery alone without MRI-based MDT discussion resulting in positive CRM in 16 cases (26%) as compared to 1 out of 116 (1%) in those patients with MDT discussion of MRI. Overall CRM+ve rate in all nonpalliative patients with or without MDT discussion was 12.5% (32 out of 256), significantly lower than the <20% rate (P<0.001) quoted in national guidelines. Re-audit in 98 consecutive patients following a change of policy produced a lower CRM+ve rate of 3% (1 out of 37) for all surgery alone patients and an overall CRM+ve rate of 7% (5 out of 70). In conclusion, MDT discussion of MRI and implementation of a preoperative treatment strategy results in significantly reduced positive CRM in rectal cancer patients.