In vivo metabolite profiles are a powerful characteristic of children's posterior fossa tumours and can be used to aid their non-invasive diagnosis. However, information is limited to a small number of detectable metabolites and it is challenging to accrue representative numbers of rare tumours. Tissue metabolite profiles can be obtained on small samples (10 to 50mg) using high resolution magic angle spinning magnetic resonance spectroscopy (HRMAS). This technique can rapidly acquire metabolite profiles providing more extensive metabolite information than available in vivo. The use of tissue banks can help to provide more samples on the rarer tumour types such as ependymoma aiding their characterisation. Metabolite profiles were obtained from freshly frozen ependymomas (n = 18), medulloblastomas (n = 37) and pilocytic astrocytomas (n = 24) using HRMAS and concentrations were quantified using the Mestrenova software package. Automated pattern recognition techniques were used to classify the samples according to tumour type. Elevated myo-inositol was characteristic of ependymoma whilst high taurine, phosphocholine and glycine distinguished medulloblastoma. Glutamine, hypotaurine and NAA were increased in pilocytic astrocytoma. Metabolite profiles gave a cross-validated diagnostic accuracy of 94%, 92% and 88% for ependymoma, medulloblastoma and pilocytic astrocytoma respectively. Metabolite profiles of tissue can classify posterior fossa tumours with enhanced metabolite information. The strength of HRMAS lies in its ability to rapidly obtain metabolite profiles from small pieces of tissue in a non-destructive manner. The tissue retains its structure and protein markers, and so is available for future immunohistochemical staining and other investigative techniques on the same piece of tissue.