Abstract: A hybrid sampling-based RRV (rapidly-exploring random vine) algorithm named HS-RRV is proposed to solve the inefficient motion planning problems of the space manipulator for on-orbit assembly in crowded and cluttered environments. Firstly, a strategy for sampling in the workspace and configuration space simultaneously is proposed, where the sampling weight is dynamically adjusted according to the algorithm's progress, and thus the workspace information is fully utilized to reduce the sampling regions in premise of the algorithm completeness, so as to improve the planning efficiency. Furthermore, the hierarchical quadratic least-square programming (HQLP) method is utilized in the local planner, where the kinematic and dynamic properties and constraints of the manipulator are considered to improve the viability of the planned trajectory. When the algorithm is trapped in complex regions, the methods of the bridge test and the principal component analysis (PCA) are used to identify the types of the local space so as to obtain more efficient expansion directions. Finally, the proposed method is compared with several known methods in multiple assembly scenarios through simulation, and the results show that the proposed method has higher computation efficiency and planning success rate, as well as a relatively shorter trajectory length.