[Objectives] In order to give consideration to the cost and accuracy of ship maneuvering motion prediction, [Methods] A fast and highly accurate ship maneuvering motion prediction method was presented based on the numerical calculation method, combined with the sensitivity analysis of hydrodynamic derivatives. Using RANS equations, VOF algorithm, and dynamic grid method, three PMM tests of the DTMB5415 ship model were simulated. The linear hydrodynamic derivatives obtained from the regression agrees well with the experimental data, which verifies the validity of the adopted CFD approach. The ship maneuver mathematical model was established based on the MMG method, and the Runge-Kutta algorithm was utilized to solve the equations to predict the model's turning circle and zigzag maneuver motion. The sensitivity of hydrodynamic derivatives of two maneuvering motions were calculated. [Results] The trajectory and parameters simulation results are in fair agreement with the experimental results. The average error of maneuver motion parameters of turning and zigzag is 5.1%, and 11.7% respectively. Compared with the results of self-propelled ship model simulation using CFD, both of the accuracy and cost are improved. The results of hydrodynamic sensitivity analysis also verified that some nonlinear hydrodynamic derivatives has a little influence on the manoeuvrability criterion, and the empirical formula can be used for estimate. [Conclusions] The results show that the proposed method can satisfy the demands of engineering application in the prediction of the maneuvering motions at the preliminary design stage. It is feasible to predict the maneuverability of ships by solving the hydrodynamic derivatives with low sensitivity by empirical formula instead of numerical calculation.