<p>The accurate estimation of absorbed doses outside the treatment field is relevant in pediatric radiotherapy due to radiosensitivity and long-life expectancy. This study aims to evaluate out-of-field doses in pediatric brain radiotherapy by comparing dose calculations from a commercial treatment planning system (TPS) with Monte Carlo (MC) simulations. A pediatric computational phantom representing a 10-year-old child treated for a brain tumor using Volumetric-Modulated Arc Therapy (VMAT) was used. Dose calculations were performed in the Eclipse TPS and compared with MC simulations carried out using PRIMO software. The MC source configuration, dose-normalization procedure, VMAT control-point reproduction and uncertainty assessment were described to improve reproducibility. Agreement between measurements and MC simulations was observed for reference PDDs, whereas discrepancies were obtained for small-field lateral profiles, mainly due to dose gradients and detector volume-averaging effects. Outside the treatment field, the TPS showed dose discrepancies when compared with MC simulations, particularly in organs located far from the target volume. In several OARs, the TPS underestimated the mean absorbed dose, while in some structures closer to the field, dose overestimation was observed. These out-of-field values were interpreted with caution because organ-specific experimental validation in the pediatric phantom was not available. Within the limitations of a single phantom and plan, MC simulations provide a complementary assessment of out-of-field doses and anatomical volumes when compared with TPS calculations. Their combined use can improve dosimetric characterization and support future pediatric radiotherapy studies, although experimental validation in anthropomorphic phantoms remains necessary before claiming patient-specific accuracy.</p>