The purpose of this dissertation is to present the development and testing of the parallelisation of a Range-Doppler SAR processor. The inherent data parallelism found in SAR data lead to the choice of using master slave parallel processor, where copies of a slave task perform the same tasks on different sets of data. However, the SAR processor that was parallelised needed to implement a corner turn without saving data to disk keeping the data set being processed distributed in memory over the nodes in the cluster. This was successfully achieved using a in-place method, thus saving valuable memory resources. Once the parallel processor was implemented some timing tests where performed, yeilding a maximum speedup factor of 6.2 for an 8 slave processor system.