Redundant manipulators with flexibility have been studied and applied in many fields. Particularly, trajectory tracking for redundant manipulators warrants further investigation. Different from these schemes presented in Chaps.  2 and 3 , this chapter presents a model predictive control (MPC)-based scheme to achieve trajectory tracking for redundant manipulators. For the nonlinear model of manipulators, linearization is performed to derive predictive outputs via the forward kinematic equation. Subsequently, an MPC scheme is formulated to minimize position error, joint velocity norm, and joint acceleration norm, while explicitly incorporating three-level joint limits and a terminal equality constraint. This scheme is further simplified into a convex quadratic programming (QP) problem. Furthermore, a zeroing-type dynamic neural network (ZDNN) model is developed to solve the constructed MPC scheme. The presented MPC scheme solved via the ZDNN model is compared with existing planning schemes and conventional solvers through both numerical simulations and physical experiments, examining performance under both disturbance-free conditions and scenarios with sudden external disturbances. Simulation results verify that the presented MPC scheme solved by the ZDNN model achieves satisfactory trajectory tracking for redundant manipulators, outperforming other existing schemes and solvers in efficiency, response speed, and robustness.