Central nervous system (CNS) tumors represent the second most prevalent group of cancers in children and adolescents, yet account for the majority of childhood cancer-related deaths and considerable morbidity among survivors, due to high-intensity non-selective standard therapies delivered to immature nervous system structures undergoing development. These tumors arise at different ages –not infrequently very early in life-, in different locations and cellular contexts, have varied cell types of origin, and have heterogeneous responses to the “classic” current therapeutic approaches. Demographic, radiologic and morphological characterization have several limitations, putting into the “classic boxes” heterogeneous tumors that are diverse in their genetic and epigenetic background and that will likely behave biologically different. Given that, epigenetic disruption (i.e. DNA methylation, histone modification and chromatin remodeling) is a common feature identified more and more frequently in pediatric cancer, it is logical to speculate that interrogating epigenetic marks may help to further define the molecular profile, and therefore tumor biology, evolution and treatment of these tumors. An integrated approach that incorporates traditional features complemented with genetic and epigenenetic specific markers offers tremendous promise to “risk-group” stratification and better prognostication. Also, it will help unveil the key driver pathways for tumor formation and for the discovery of targeted therapy for neoplasms that appear in the developing brain, facilitating early identification of therapy responders and track accurately disease progression. In this paper, we reviewed the most representative pediatric brain tumors where epigenetic alterations have been identified as initiating or driving events in tumor development, maintenance or progression.