The surface of Mars is a complex terrain covered with soft soil， which poses significant risks of subsidence， high slip， or even collapse for rovers exploring the planet. A thorough analysis of rover trafficability and effective path planning are crucial in mitigating these risks. This paper discusses a comprehensive review of the strategies and methods employed by the Sojourner， Spirit， Opportunity， Curiosity， Perseverance， and Zhurong rovers and the latest advances in trafficability assessment based on multi-information fusion， deep learning， and data-driven techniques. Finally， this paper outlines the prospects for advancing the field of rover trafficability assessment.