The incidence of brain tumors among children is second only to acute lymphoblastic leukemia, but the mortality rate of brain tumors has exceeded that of leukemia, making it the most common cause of death among children. Medulloblastoma (MB) is the most common type of brain tumor among children. Malignant brain tumors have strong invasion and metastasis capabilities, can spread through cerebrospinal fluid, and have a high mortality rate. In 2010, the World Health Organization first divided MB into four molecular subtypes based on molecular markers: WNT, Sonic hedgehog (SHH), Group 3, and Group 4. MB is a highly heterogeneous tumor. Different molecular subtypes of MB have significantly different clinical, pathological, and molecular characteristics. The prognosis of MB varies significantly among patients with different subtypes of this cancer. Thus, it is needed to study new diagnostic and therapeutic strategies. Metabolomics is an advanced analytical technology that uses various spectroscopic, electrochemical, and data analysis technologies to study and analyze the body’s metabolites. By detecting changes in metabolite types and quantities in different types of samples, it can sensitively discover the physiological and pathological changes in the body. It has great potential for clinical application and personalized medicine. It is promising and can help develop personalized treatment strategies based on the metabolic profiles of individuals. It can unravel the unique metabolic profiles of MB, which may revolutionize our understanding of the disease and improve patients’ outcomes.