Cancer metabolism is an essential aspect of tumorogenesis, as cancer cells have increased energy requirements in comparison to normal cells. Metabolomic techniques can provide quantitative data for a large number of small molecules in tissues and enable the analysis of multiple intricate metabolic pathways. Positron emission tomography (PET) using 18F-Fluorodeoxyglucose (FDG) enables in vivo analysis of glycolysis and is widely used in oncology. High tumor FDG uptake is a prognostic factor in breast cancer and has been associated with tumor aggressively. Seventy breast cancer samples obtained from untreated patients who had underwent FDG-PET imagery were analyzed through an untargeted metabolomic approach using liquid chromatography-mass spectroscopy (LC-MS) to study possible correlations between metabolomic data and FDG uptake. Tumors were split into two groups depending on avidity for FDG as measured with PET. The Compound Discoverer 4.0 software enabled identification of 854 metabolites. PLSDA based models predicted FDG uptake with an accuracy ranging from 0,73 to 0,77. Selected metabolites varied depending on the use of scaling or log transformation. Metabolites correlated with tumor FDG uptake were, among others, glutathione, amino-acids such as glutamate, proline or tyrosine, L-acetyl-carnitine, metabolites from the kynurenine pathway such as L-kynurenine or formyl-kynurenine and polyamines such as N1,N12-diacetylspermine or N1-acetylspermine. These metabolites have been previously shown to reflect cancer aggressivity. The correlation between the glycolytic pathway activation and tumor FDG uptake could not be directly assessed but indirect signs showed a higher glycolytic activity in tumours presenting a higher FDG uptake. Studying new metabolites identified through this process could enable a better understanding of tumor metabolism and identification of new biomarkers.