For an uncertain manipulator systems with unknown disturbances and actuator input saturation, a fixed-time sliding mode control method based on an neural network observer is proposed. First, considering the model parameter uncertainty and unknown disturbances, the system is approximated by an RBF neural network, and a fixed-time neural network observer is designed to quickly estimate the unknown joint position and velocity information, so that the system estimation error can converge to zero within a fixed time. Furthermore, based on the proposed saturation compensator scheme, a novel fixed-time sliding mode controller for the closed-loop system is designed, which not only achieves the fixed-time reachability of the sliding mode surface but also guarantees that the position tracking error signal converges to zero within a fixed time. Finally, the simulation experiment based on a two-joint manipulator system, verifies that the control scheme designed improves the tracking control performance of the manipulator.