Abstract AIMS Glioblastoma (GB) is an inherently heterogenous and invasive primary brain tumour. Genetic and transcriptomic studies have attempted to classify GB into subtypes that can identify therapeutic vulnerabilities and predict survival. An outstanding question and technical challenge is to visualise the metabolism of a tumour within its native microenvironment and understand to what extent its metabolism is driven by the microenvironment. METHOD Patients with GB tumours were infused with [U-13C]glucose and intra-operative multi-regional tumour samples were rapidly snap frozen for metabolic tracing analysis using mass spectrometry imaging (MSI). Tissue microenvironment was assessed using multiplexing of antibodies on immunocytochemistry (IMC) of parallel sections. Isotope tracing in human derived neurospheres and patient derived xenografts were used to confirm the presence of metabolic phenotypes detected in the human samples and to test susceptibility of each phenotype to a panel of drugs targeting cellular metabolism. RESULTS Three metabolic signatures were identified in patient GB tumours: glycolytic, oxidative and a mixed glycolytic/oxidative phenotype. The phenotypes did not correlate with microenvironmental features, including proliferation rate, immune cell infiltration, and vascularisation. The phenotypes were retained when patient derived cells were grown in vitro, or as orthotopically implanted xenografts, and were robust to changes in oxygen concentration, demonstrating their cell intrinsic nature. The spatial extent of the regions occupied by cells displaying these distinct metabolic phenotypes are large enough to be detected using clinically applicable metabolic imaging techniques. The phenotypes also demonstrate differential drug sensitivities. CONCLUSION This is the first report to demonstrate high resolution imaging of metabolic fluxes in a human tumour in vivo. In conjunction with studies on patient-derived neurospheres and orthotopically implanted xenografts we have demonstrated the presence of different metabolic phenotypes within GB that are tumour cell intrinsic and largely independent of the tumour microenvironment. These can be imaged clinically and used to stratify patients for personalised therapy.