Preventing collisions at sea is one of the main tasks of the safe ship navigation, attracting great attention. Such artificial intelligence methods as neural networks and fuzzy logic have been applied to the problem. The paper aims to address the collision avoidance issue by using the dynamic predictive guidance technique. The Visual C++-based simulation framework has been developed including the ship's nonlinear dynamic model and the algorithm of numerical optimization. The case study of the predictive guidance system was carried out in a two-ship crossing situation. The simulation results show the feasibility and effectiveness of the proposed technique and software.