BACKGROUND: The treatment for childhood intracranial ependymoma includes maximal surgical resection followed by involved-field radiotherapy, commonly in the form of intensity-modulated radiation therapy (IMRT). Proton-beam radiation therapy (PRT) is used at some centers in an effort to decrease long-term toxicity. Although protons have the theoretical advantage of a minimal exit dose to the surrounding uninvolved brain tissue, it is unknown whether they have the same efficacy as photons in preventing local recurrence. METHODS: A retrospective review of medical records from September 2000 to April 2013 was performed. Seventy-nine children with newly diagnosed localized intracranial ependymomas treated with either IMRT (n 5 38) or PRT (n 5 41) were identified, and progression-free survival (PFS) was analyzed with Kaplan-Meier and Cox multivariate analyses. RESULTS: The median age at diagnosis was 3.7 years for all patients (range, 0.4-18.7 years). There were 54 patients with infratentorial tumors (68% of the total population). Patients treated with PRT were younger (median age, 2.5 vs 5.7 years; P 5.001) and had a shorter median follow-up (2.6 vs 4.9 years; P <.0001). Gross total resection (GTR) was achieved in 67 patients (85%) and was more frequent in the PRT group versus the IMRT group (93% vs 76%; P 5.043). The 3-year PFS rates were 60% and 82% with IMRT and PRT, respectively (P 5.031). CONCLU-SIONS: Children with localized ependymomas treated with PRT have a 3-year PFS rate comparable to that of children treated with IMRT. This analysis suggests that local control is not compromised by the use of PRT. The data also support GTR as the only prognostic factor for PFS.