PURPOSE: We analysed the cohort of paediatric patients with metastatic non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated in the BERNIE protocol, i.e. open-label, multicentre, randomised phase II study evaluating the role of bevacizumab (BO20924/ITCC-006; ClinicalTrials.gov: NCT00643565). METHODS: Eligible patients were randomised 1:1 to add or not add bevacizumab to nine courses of intensive multi-drug chemotherapy, followed by 12-month maintenance chemotherapy (plus surgery and radiotherapy). The primary end-point was event-free survival (EFS); secondary objectives were objective response rate (ORR) and overall survival (OS). RESULTS: From 2008 and 2013, 49 NRSTS patients (out of 154 cases) were treated, 26 in the standard arm and 23 in the bevacizumab arm. ORR was seen in 10 out of 36 evaluable cases (27.7, i.e. 4/18 standard arm cases and 6/18 bevacizumab arm cases. Two-year EFS was 27.395CI] 13.9-42.5) for all NRSTS patients, i.e. 34.9954.6-56.2) for bevacizumab arm and 22.995.1-43.9) for standard arm (p-value = 0.19). Three-year OS (median follow-up 48.6 months) was 35.2 with no differences in the two arms. Time to event and time to death were 16.3 and 17.2 months for bevacizumab arm and 2.1 and 7.6 months for standard arm, respectively. Patients not receiving any local treatment on primary disease had a worse outcome as compared to others. Treatment results were better for patients receiving surgical resection and worse for those who did not receive any specific treatment. CONCLUSION: The addition of the anti-angiogenic agent to the standard chemotherapy did not show statistically significant improvement in survival in metastatic NRSTS.