This paper investigates the fault-tolerant control problem for a class of continuous-time Markovian jump systems with actuator faults, nonlinearity and external disturbances. In this study, the exact knowledge for actuator fault and model nonlinearity is totally unknown, and the considered external disturbance contains matched and mismatched part which is more general and widely exists in practical engineering systems. Since sliding mode control exhibits limitations to be employed to stabilize jump systems (Xia et al. 2009) in this paper, an adaptive control scheme with backstepping method is developed to solve this control problem. By on-line estimating the loss of effectiveness of actuators faults and the bounds of Euclidean norm of nonlinearity function, the proposed adaptive backstepping controller can compensate these effects completely, and the stochastic stability of the fault closed-loop system is guaranteed. A numerical example with a wheeled mobile manipulator model is included to verify the effectiveness and show the potential of the proposed new design techniques. (C) 2015 Elsevier Ltd. All rights reserved.