Medulloblastoma is the most common paediatric brain cancer. Standard therapies have not changed in decades, causing survival outcomes for children with aggressive medulloblastoma to stagnate. Immunotherapy has recently become a major focus of novel treatment development, and while there are multiple clinical trials aiming to increase immunological recognition of medulloblastoma, none have demonstrated efficacy to date. To tailor future immunotherapies to medulloblastoma, a deeper understanding of the interactions between medulloblastoma and immune cells is crucial. Using a murine allograft model of medulloblastoma in both immune-competent and immune-deficient mice, we show that medulloblastoma growth, and response to clinical treatments, is independent of the adaptive immune system. Furthermore, standard-of-care craniospinal irradiation (CSI) depletes adaptive immune populations in the brain, and enhances brain-resident microglia and peripheral macrophage infiltration into the medulloblastoma microenvironment. This revealed an opportunity where immunotherapies targeting myeloid cells might have potential to be combined with CSI in a clinical setting. Magrolimab (Hu5F9-G4) is a humanised antibody that targets CD47 with high affinity, however, we show this treatment is ineffective as a single agent in a patient-derived xenograft model of Group 3 medulloblastoma. Using a small animal radiotherapy platform, we have developed preclinical CSI protocols that mimic clinical treatment responses, where radiotherapy causes temporary medulloblastoma regression followed by disease regrowth. Remarkably, when CSI was combined with magrolimab we observed marked and persistent tumour regression in mice. Our work is currently employing a range of techniques, including single cell sequencing, to elucidate the mechanisms by which radiotherapy alters the medulloblastoma microenvironment to enhance the anti-tumour activity of innate immune cells in the brain. Importantly, medulloblastoma is a paediatric disease. To enhance our preclinical evaluation of CSI and magrolimab, we have developed new mouse models that more accurately reflect the immune system of children and show that the innate immune populations in paediatric brain are distinct from adult mouse brain. By evaluating this novel combination of immunotherapy with standard medulloblastoma treatments, in age-appropriate models our research should facilitate the rational selection and rapid translation of optimised treatment combinations for future medulloblastoma clinical trials.