For recent research on shipboard computer systems to support the safety navigation in rough seas, potential risk estimate is one of the most important problems to be solved. In order to keep the ship on the safe side, potential risk estimate and prediction are indispensable functions to be offered to the navigator. With these two functions, the system will be able to support ship's operation in rough seas. To realize them on shipboard computer system, probabilistic method is useful. In this report, a proposal is made for probabilistic potential risk estimate on longitudinal strength. On the estimate, wave-induced bending moment, slamming-induced whipping moment and time to escape from rough weather are took into account by use of Ferro's theory. As an example, this method is applied to full scale data of a container ship, Japan Alliance, operating in the North Pacific service.