Altered metabolism is one of the hallmarks of cancer and is known to enable the proliferation and survival of cancer cells in changing microenvironments. Metabolic changes can be interrogated using non-invasive metabolic imaging techniques, which enable the investigation of tumour metabolism in vivo and the identification of imaging biomarkers. Hyperpolarised 13C Magnetic Resonance Spectroscopic Imaging (MRSI) is a metabolic imaging technique that enables dynamic imaging of isotopically labelled metabolites in vivo, with [1-13C]pyruvate being the most widely used substrate in pre-clinical studies and the only substrate used so far in clinical studies. In the first part of this thesis, hyperpolarised 13C pyruvate MRSI was directly compared to mass spectrometry imaging in a murine lymphoma model, cross-validating both methods for providing information on the relative spatial distribution of 13C-labelled lactate. In the second part, the potential of hyperpolarised [1-13C]Ketoisocaproic Acid as a substrate for hyperpolarised imaging was investigated. Ketoisocaproic acid is a substrate for Branched Chain Aminotransferase 1 (BCAT1), the first enzyme in the catabolic pathway of Branched Chain Amino Acids. This enzyme is upregulated in many cancers and is linked to cancer progression. In this project, the role of BCAT1 in cell proliferation was confirmed in glioblastoma patient derived cells and tumours and BCAT1 was shown to regulate the cells’ metabolic and transcriptional profiles through regulation of alpha-ketoglutarate metabolism. In vivo imaging of BCAT1 activity is therefore suggested as a potential technique for patient stratification and potentially treatment monitoring. [1-13C]Ketoisocaproic acid was successfully hyperpolarised and showed fast uptake into subcutaneous EL4 lymphoma tumours, healthy rat brain and orthotopic gliomas in rat brain, following intravenous injection. Sufficient 13C leucine signal was generated in the lymphoma tumour models enabling in vivo imaging of BCAT activity but the reaction was found to be significantly slower in the healthy rat brain and in the glioma tumours. This was explained by low levels of glutamate, a co-substrate for the reaction, in the glioma tumours, which is a potential limiting factor for this reaction.