570 Background: 20% of rectal cancer patients have metastatic lymph nodes outside the mesorectum (EMRs). These EMR node positives are associated with poor prognosis. Accurate selection would help to tailor treatment and improve prognosis for these patients. Methods: Rectal cancer patients were included in a study in which treatment was based on (contrast-enhanced) MRI. EMR-status was predicted by an expert radiologist. Based on this prediction patients underwent chemoradiation (CRT) of the EMRs. 6-8 weeks after CRT EMRs were restaged. If still involved, the EMRs were resected. When the EMRs were sterilised by the CRT, they were not resected. Patients were followed 3 to 6-monthly after surgery by a combination of modalities. 3-year outcome was estimated with Kaplan-Meier curves. Results: 50 patients with suspected EMRs were included. Median follow-up was 26(0-50) months. In 13 patients EMRs were resected after CRT and in only 2/13 positive nodes were found. Of the remaining 37 patients 5 had a local recurrence (LR). In total 32+11=43 patients (86%) had no involved EMRs after CRT. Five patients had metastasis of whom 2 also had a LR. 3-year LR was 2.3%, 3-year DFS was 86% and 3-year OS was 90%. Conclusions: MRI-based selection of patients with EMRs for CRT provides adequate local control. Distant metastasis is the main cause of poor prognosis in these patients. When MRI is used for EMR identification and thus for identification of patients who need CRT also on the obturator regions, patients can be spared an extensive resection with associated morbidity.