The trajectory tracking control of a SCARA industrial robot in joint space was studied. In general, system unmodeled dynamics, joint friction and gap, and unknown load will lead to change of the robot dynamic performance, which will affect the robot trajectory tracking performance, and so does the external disturbances.Aiming at this issue, a fast nonsingular terminal sliding mode control strategy based on cuckoo search algorithm was proposed.In this method,the optimal mechanism of cuckoo search algorithm was used for trajectory planning.Under the framework of Lyapunov stability theory, the fast nonsingular terminal sliding mode surface was adopted to compensate and restrain the unmodeled dynamics and external disturbances, and the fast terminal sliding mode was introduced to improve the response speed. Meanwhile, an Anti-Windup technique was proposed to compensate the nonlinear factors of the deadzone in the system.According to Lyapunov stability theory, it is proved that the system trajectory tracking error is global stable. The effectiveness of the proposed approach is tested through a trajectory tracking experiment.