<p>Tumors of the central nervous system (CNS) are a&#xa0;diverse group of neoplasms in the brain or spinal cord that commonly occur in children and adolescents. For decades, they have been primarily classified based on their histological appearance and immunohistochemical characteristics; however, traditional neuropathological techniques have reached their limits in capturing the full spectrum of biological heterogeneity of these tumors. The discovery of distinctly identifiable types of pediatric brain tumors based on the distribution pattern of methyl groups on the tumor DNA laid the foundation for a&#xa0;novel tumor classification that is independent of their morphology and utilizes AI-assisted algorithms for diagnosis. The advantages lie in its objectivity and reproducibility as well as relatively low material requirements and costs. The establishment of the DNA methylation-based classification of CNS tumors has revolutionized neuropathological practice over the past decade and was recognized by the World Health Organization in 2021 as an important diagnostic criterion. The fact that German health insurance companies now cover the costs of this method as a&#xa0;reference service for pediatric CNS tumors is an outstanding example of the rapid translation of technical innovation into direct patient care. Future challenges lie in using these diagnostic advances for improved design of clinical trials and the development of more effective treatment strategies, applying them to other types of tumors, and comprehensively disseminating them worldwide.</p>