This thesis presents a new methodology for high precision nanoscale machining using Focused Ion Beam (FIB) processes. The methodology is supported by several novel models and methods developed during the PhD project. Gallium focused ion beam instruments are capable of processing virtually any material with a nanometre resolution. This has established FIB based instruments as invaluable specimen preparation tools in material science and circuit editing and failure analysis tools in the semiconductor industry. So far, the technique has had limited application in nano and micro- manufacturing, due to the high cost of the equipment and the long process cycle times required . Nonetheless in recent years it has been demonstrated that FIB can be a viable manufacturing technology if employed in the fabrication of high precision replication tools and it has the potential to replace existing electron and photon lithography techniques. One of the current problems is that the existing FIB procedures developed for material science are often not optimised for quality and efficiency or not applicable in manufacturing. A new machining methodology has been proposed that can be used as a guide to optimise FIB processes for improved efficiency and production quality. The methodology systematically looks into the material selection, the choice of gas precursor and the optimisation of the scanning parameters. To accomplish this several new models and methods are developed. A raster scanning model is proposed that links the probe current, the dwell time, the number of loops and the step with the key process parameters of refresh time, exposure time, dose, and dose distribution. Furthermore, a new term apparent beam size and a method for its measurement are suggested as an alternative to the commonly used "knife edge diameter". The apparent beam size is found to be material and precursor dependent and together with the overlap is accounted for as a key factor in the dose uniformity criterion formulated in the project.