Abstract: Objectives To address the problems of low search efficiency, large deviations in terminal position and attitude, and the difficulty in balancing safety and feasibility in underactuated ship path planning, a ship berthing path planning method based on improved bidirectional hybrid A* algorithm is proposed. Methods By incorporating ship motion constraints and speed variations across different berthing stages, an improved bi-directional hybrid A* search mechanism (IBHA*) is introduced to improve search efficiency while ensuring terminal position and attitude constraints. Meanwhile, a path search cost function incorporating path length cost, steering cost, and safety distance cost is constructed to enhance navigation safety and sailing efficiency. In addition, Bezier curves with curvature constraints are employed to smooth the generated berthing paths, thereby further improving path smoothness and reachability. Results Experimental results demonstrate that the proposed berthing path planning algorithm reduces the average berthing path length by 6.12% and the average path search time by 36.56% across various scenarios, which is capable of generating berthing paths that achieve a balance among navigation safety, berthing stability, and planning efficiency. Conclusions The proposed IBHA* ship berthing path planning method effectively balances search efficiency and path feasibility, and achieves a coordinated optimization of navigation safety, stability, and berthing efficiency. It provides a reliable and practical path planning solution for autonomous ship berthing.