A quantitative measure of myocardial perfusion is of use clinically to provide information about the severity and location of a perfusion abnormality. Thus it can help in guiding the clinician in the choice of therapeutic intervention, and in monitoring changes in the perfusion abnormality following intervention. The current methods of measuring perfusion have been exmained and this thesis has assessed whether MRI could become a useful tool. MRI is non-invasive, and fast imaging sequences have opened the possibility of dynamic imaging. This, coupled with high spatial resolution, allows assessment of the transmural layers of the myocardium. In MR imaging, a measure of absolute myocardial perfusion (ml/min/g) is not possible at present because the MR signal is not directly related to the tissue concentration of the contrast agent. The concentration of the bolus needs to be accurately tracked through the tissue, in order for tracer kinetic models, which enable perfusion to be calculated, to be of use. This thesis has tried to cover the difficulties of using MRI for quantitatively assessing myocardial perfusion and has tried to correct the errors which are introduced by the simplistic use of MRI. In this way, the imaging methodology has been optimised. The optimised imaging techniques have been applied to patients with acute myocardial infarction. These patients were dynamically imaged using a new contrast agent Gd-BOPTA. The concentration of the agent was estimated, and a tracer kinetic model applied. The values estimated for perfusion were higher than expected, and brought into question the behaviour of the contrast agent. The final experimental section has investigated whether Gd-BOPTA may behave differently to currently available contrast agents (e.g. Gd-DTPA), and so be responsible for the spurious perfusion results. Data are shown which support this argument, and further work is required in applying this optimised imaging technique of perfusion measurement using Gd-DTPA.