In this paper, a fixed-time repetitive learning control scheme is proposed for uncertain rigid robot manipulators. Different from the existing fixed-time control schemes, a simple nonsingular fixed-time virtual controller is constructed to directly avoid the singularity caused by the differentiation of the virtual controller. Then, a robust control law is developed to guarantee the effective compensation of the left non-periodic uncertainty. With the proposed control scheme, the fixed-time error convergence in the transient process and high precision tracking performance in the steady-state process can be both guaranteed simultaneously. Simulation results on a two-link robot manipulator verify the validity of the proposed method.