This thesis reports on an investigation into the simulation of ship manoeuvring in restricted waters; in particular ship squat and ship-bank interaction. A time-domain mathematical model has been developed to predict unsteady squat and dynamic acceleration effects for a vessel travelling in non-uniform water depth. A quasi-steady approach was adopted, where the prediction at each time step is based on steady-state heave force and pitch moment in uniform water depth. A comprehensive set of model scale experiments was conducted to investigate squat in both uniform and non-uniform water depth. Regression analysis was performed on the uniform water depth test data to develop empirical equations for steady-state squat, which were used as input to the mathematical model. The equations apply to a full form vessel and are dependent on under keel clearance parameters, channel width parameters and depth Froude number. Empirical correction factors were also developed to estimate the effect of propulsion on squat. Good correlation was observed between predictions from the squat mathematical model and steady-state sinkage measurements for a wide range of water depth to draught ratios. Predictions from the mathematical model were also compared with unsteady sinkage measurements on a ship model travelling over a simplified ramp bank. The general trend of the predicted unsteady sinkage was reasonable, improving the realism of simulation for abrupt changes in water depth compared to predictions where acceleration effects in the vertical plane are neglected. However, the maximum unsteady sinkage was not always predicted accurately, which may be attributable to limitations and assumptions associated with the technique. A comprehensive model scale experiment program was conducted to investigate ship-bank interaction. The following parameters were systematically varied for three different hull forms: flooded bank height, water depth, bank slope, lateral ship to bank distance, vessel speed and vessel draught. The results from these experiments were used to develop a bank parameter to estimate the effects of lateral surface piercing and flooded banks. This parameter was utilised in regression analyses to develop empirical formulae for steady-state bank induced sway force and yaw moment. The empirical formulae were validated against independent model scale measurements from literature and showed reasonable agreement for a range of cases. The formulae were then incorporated into the existing mathematical model of the AMC ship-handling simulator. The quasi-steady technique was used to predict the path of a vessel for a real-life manoeuvre where bank induced yaw moment is used to aid a turn. The proposed bank effect simulation method was found to provide a satisfactory solution to the problem of ship-bank interaction simulation.