Standard surgical training consists of the acquisition of theoretical knowledge complemented with practical observation during surgery and, at some advanced stage, performing the surgical procedures on live patients under the supervision of qualified surgeon. Also, experienced surgeons still require training to perform rare and complex cases. However, a potentially contradictory situation exists whereby patients' expectations of surgical experience are obviously high, whereas clinical governance requires surgeons' track records to be established before they move to operate on live patients. There is increasing fear of litigation among healthcare professionals, along with worktime legislation which limits the hours available for training. Therefore there is a need for surgical simulation systems for training to overcome the initial lack of experience and for planning of rare and complex procedures. In this thesis we have developed a Cataract Eye Surgery Simulation system (CESS) with the capability to simulate, in addition to the basic surgical interaction, the main steps of cataract eye surgery, and which can be used as a teaching and training method to train medical students in a realistic environment. The CESS, in its present form, provides users with an interactive, affordable, easy to learn, risk-free, reusable and adaptable means for training in cataract eye surgery procedures. In this thesis we have presented and examined different modelling techniques that have been used for soft tissue simulations. In particular, we have been interested in the computation complexity of the models, numerical stability, and physical realism. Within this context, we proposed the Circular Wave Model (CWM) which is capable of modelling homogeneous, surface objects and is fast enough for applications such as surgery simulations. The CWM algorithm is derived using the analogy of water waves produced when water is hitby stone. Two dimensional (2D) and three dimensional (3D) objects can interactively be deformed with the CWM. We have put forward a new technique to assign stiffness and damping coefficients to springs of the mesh being manipulated and the spring that connects the virtual surgical instrument to its manipulated mesh.