In order to improve the motion accuracy of the 5-DOF manipulator, the motion control of the 5-DOF manipulator was researched based on the global optimal sliding mode controller. The mechanical structure of the 5-DOF manipulator was analyzed, and its structure composition and motion principle were clarified. Through the rotation angle of each joint node and the length of the connecting arm, the rotation relationship between the adjacent joint nodes and the movement vector of the joint node relative to the basic coordinate were obtained respectively, and the dynamic equation of the manipulator was obtained through the motion vector. The system equation of a single joint node was calculated through the dynamic equation of the manipulator. The sliding surface was constructed by using the angle error, and the sliding mode controller was established by using the system equation of a single joint node. In order to obtain the optimal solution of the sliding mode weighting factor and the reaching weighting factor in the sliding mode controller, the particle swarm optimization algorithm was used to search for the global optimal sliding mode controller in the solution space, which was used to control the motion process of the 5-DOF manipulator. In the experiment, the motion process of the joint node and the end effector were controlled by the proposed method and the disturbance observer method. From the control results, the control accuracy of the proposed method is 1712% and 1349% higher than that of the disturbance observer method respectively. It shows that the proposed method can be used to guarantee motion accuracy of the 5-DOF manipulator effectively.