Magnetic Resonance Spectroscopy (MRS) for the evaluation of in vivo cerebral metabolite ratios is a relatively new radiological modality, which permits the detection and evaluation of specific metabolites within the human body. As with all imaging modalities, the accuracy of the equipment to perform its given task is paramount and the effect of introduction of elements and/or contrast agents to a study, must be understood before a diagnosis or prognosis can be made with any degree of certainty. The following chapters describe the development and testing of a phantom, the development and testing of metabolite simulating solutions and the experiments conducted to determine the spatial localisation accuracy of a MRS imaging system. This thesis begins with a description of the steps undertaken to investigate a clinical MRS system's capacity to reliably produce spectra over both the long and short term. Validation of the phantom construction, metabolite simulating solutions and spatial localisation accuracy is presented in detail. The construction of the localisation phantom resulted in a two compartment perspex phantom and the development of the aqueous metabolite solutions produced two solutions with distinct and separate metabolite peaks with comparable peak amplitudes and sufficient line width separation to prevent possible metabolite resonance cross contamination but without illustrating any significant susceptibility artefacts. Development of the metabolite simulating solutions was a major part of the work conducted. Short and long term (1 year) reproducibility of the measured metabolite peak amplitudes were assessed for four different pulse sequences. Of these, the most reproducible results were obtained with a TE 270ms PRESS sequence (coefficient of variation < 0.6% (short term) and < 3% (long term)). The spatial localisation experiments illustrated a disturbing error in the placement of the user prescribed volume of interest (VOI) with respect to the actual VOI acquisition with a difference between the two volumes of interest of up to 3 mm or 15% of the VOI size in two axes. Possible effects of magnetic resonance imaging contrast agents, specifically Gd-DTPA were also investigated. Fifty patients were included in this study. The metabolic peak amplitudes were measured pre and post-administration of the contrast agent for each subject. A paired t-test demonstrated that there was no significant difference between the mean peak heights pre- and post- administration (0.07 < p < 0.9 for the various peak amplitudes measured).