In this research article, the finite-time asynchronous sliding mode control (SMC) scheme for Markovian jump systems (MJSs) subject to sensor and actuator faulty signals, is investigated based on the average dwell time approach. Firstly, an asynchronous stochastic hybrid model is proposed in the light of the existing non-synchronization phenomenon of unmeasurable state and indirect access to jump information of original systems. The average dwell time technique, which has the ability to compensate the effects of arbitrary switching by generating a sequence of signals to regulate/choose an appropriate feedback Markov switching signal among the Markov chains, is developed during the reaching phase and sliding motion phase of the sliding mode dynamics. Secondly, based on the asynchronous stochastic hybrid model, a mode-dependent sliding mode surface function is designed. Moreover, the SMC law is synthesized such that the system state trajectories can be driven onto the specified sliding surface in a prescribed limited time. Thirdly, the finite-time analysis method combining with a mode-dependent Lyapunov function, which can reduce drastically the design conservatism, is adopted to guarantee the finite-time boundedness of the sliding mode dynamics both in the reaching phase and sliding motion phase. Fourthly, sufficient conditions are interpreted for the solutions of asynchronous controller gain matrices and an algorithm is provided to assist computing the SMC gain as an auxiliary tool. Finally, a numerical example is given to illustrate the effectiveness of the proposed new design techniques.