To remain competitive, high-quality high-value manufacturers constantly seek to improve their product quality and to reduce cost and lead times by producing ‘right first time’ machined components. Producing the right profile in such parts increasingly depends on specialised CAM packages for defining appropriate cutting strategies and tool paths. Despite the significant developments in NC simulation and verification there are still gaps between the theoretically predicted surfaces and the measured surfaces due to a relatively small percentage of machining error types that could be detected and interpreted by the existing software systems. The paper reports on a virtual environment for simulation and prediction of the deflection of thin-walled parts during machining. It aims to increase the understanding of the causes of poor geometric accuracy by considering the impact of the machining forces on the deflection of thin-walled structures. The proposed approach is based on identifying and modelling of key processing characteristics that influence part deflection, modelling of the material removal process using “voxels”, and predicting and correcting the deflection of the parts. The decision making is based on a novel integrated environment using CAD and finite element analysis tools. The reported work is part of an ongoing research on developing an adaptive machining planning environment for modelling, prediction, and selection of process and tool path parameters for rapid machining of complex low-rigidity high-accuracy parts.