Abstract: Due to the tumbling motion of the target, the dual-arm space robot to grasp a dynamic target is more challenging compared to a static target. Besides, optimization of the grasping strategy can improve the manipulability of the space robot to operate on the tumbling targets to ensure the success of the grasping mission. In this paper, a method for grasping strategy optimization is proposed based on the manipulability evaluation. When the dual-arm space robot cooperatively grasping a target, the dual-arm end-effectors contact the target simultaneously and a closed kinematics chain is formed. As a result, the formation of the closed-chain constraint complicates the evaluation of the manipulability of the space robot. First, the kinematics and dynamics of a dual-arm space robot manipulating a target are analyzed in this paper. Following this, a cooperative workspace considering the closed-chain constraint is established, and a task compatibility based detumbling manipulability metric is analyzed. The established cooperative workspace contains both the position and the attitude information of the manipulated target in the task space, which can be used for the calculation of dexterity. Then, the optimal grasping points of the end-effectors to grasping a target are determined based on the global dexterity metric, as well as the optimal grasping configuration of the space robot to grasp a tumbling target is found based on the force task compatibility metric considering the field-of-view constraint of the camera and the velocity tracking constraint of the end-effectors to the tumbling motion of the target. Using the manipulability metrics to determine the grasping strategy can make full use of the coordination of both arms to increase the manipulability of the space robot to manipulate dynamic targets, and simulations are conducted using a 7 degree of freedom dual-arm space robot to verify the feasibility and effectiveness of the proposed grasping strategy.