This paper develops a new strategy for robust fault tolerant control (FTC) of robot manipulators using an adaptive fuzzy integral sliding mode control and a disturbance observer (DO). First, an integral sliding mode control (ISMC) is developed for the FTC system. The major features of the approach are discussed. Then, to enhance performance of the system, a fuzzy logic system (FLS) approximation and a DO are introduced to approximate the unknown nonlinear terms, which include the model uncertainty and fault components, and estimates the compounded disturbance,respectively,and then integrated into the ISMC. Next, a switching term based on an adaptive twolayer super-twisting algorithm is designed to compensate thedisturbanceestimatederrorandguaranteestabilityand convergence of the whole system. The nominal controller of the ISMC is reconstructed using backstepping control technique to achieve the stability for the nominal system based on Lyapunov criteria. The computer simulation results demonstrate the effectiveness of the proposed approach.