This article provides a comprehensive review of path planning for autonomous ground vehicles in unstructured environments. It categorizes path planning into hierarchical and end-to-end approaches, emphasizing the unique aspects compared to structured path planning, such as terrain traversability analysis (TTA), cost estimation, and terrain-dependent constraints. The review discusses the factors influencing TTA, including terrain geometrical and physical properties, vehicle dynamics, and vehicle-terrain interactions. It also summarizes methods for TTA, including empirical, physics-based, and semi-empirical models, and highlights the importance of accurate TTA for path planning. The article further examines the estimation of safety, energy, and comfort costs, considering factors such as vehicle load, terrain conditions, and vehicle dynamics. It reviews the constraints imposed by terrain and vehicles, including position, continuity, kinematic, dynamic, and safety constraints. The review also covers global and local path planning methods suitable for unstructured environments, as well as end-to-end path planning approaches that rely on learning-based algorithms. The article concludes by identifying key areas requiring further research, including handling multiple uncertainties, improving the reliability and coverage of TTA, evaluating vehicle instability, considering the effect of vehicle posture on path planning, ensuring safety redundancy, and meeting adaptability and real-time requirements in path planning.This article provides a comprehensive review of path planning for autonomous ground vehicles in unstructured environments. It categorizes path planning into hierarchical and end-to-end approaches, emphasizing the unique aspects compared to structured path planning, such as terrain traversability analysis (TTA), cost estimation, and terrain-dependent constraints. The review discusses the factors influencing TTA, including terrain geometrical and physical properties, vehicle dynamics, and vehicle-terrain interactions. It also summarizes methods for TTA, including empirical, physics-based, and semi-empirical models, and highlights the importance of accurate TTA for path planning. The article further examines the estimation of safety, energy, and comfort costs, considering factors such as vehicle load, terrain conditions, and vehicle dynamics. It reviews the constraints imposed by terrain and vehicles, including position, continuity, kinematic, dynamic, and safety constraints. The review also covers global and local path planning methods suitable for unstructured environments, as well as end-to-end path planning approaches that rely on learning-based algorithms. The article concludes by identifying key areas requiring further research, including handling multiple uncertainties, improving the reliability and coverage of TTA, evaluating vehicle instability, considering the effect of vehicle posture on path planning, ensuring safety redundancy, and meeting adaptability and real-time requirements in path planning.