Tumors of the Central Nervous System (CNS) represent the leading cause of cancer-related deaths in children. Current treatment options are not curative for most malignant histologies, and intense preclinical and clinical research is needed to develop more effective therapeutic interventions against these tumors, most of which meet the FDA definition for orphan diseases. Increased attention is being paid to the repositioning of already-approved drugs for new anticancer indications as a fast-tracking strategy for identifying new and more effective therapies. Two pediatric CNS tumors, posterior fossa ependymoma (EPN-PF) type A and diffuse midline glioma (DMG) H3K27-altered, share loss of H3K27 trimethylation as a common epigenetic hallmark and display early onset and poor prognosis. These features suggest a potentially common druggable vulnerability. Successful treatment of these CNS tumors raises several challenges due to the location of tumors, chemoresistance, drug blood-brain barrier penetration, and the likelihood of adverse side effects. Recently, increasing evidence demonstrates intense interactions between tumor cell subpopulations and supportive tumor microenvironments (TMEs) including nerve, metabolic, and inflammatory TMEs. These findings suggest the use of drugs, and/or multi-drug combinations, that attack both tumor cells and the TME simultaneously. In this work, we present an overview of the existing evidence concerning the most preclinically validated noncancer drugs with antineoplastic activity. These drugs belong to four pharmacotherapeutic classes: antiparasitic, neuroactive, metabolic, and anti-inflammatory. Preclinical evidence and undergoing clinical trials in patients with brain tumors, with special emphasis on pediatric EPN-PF and DMG, are summarized and critically discussed.