A quantitative evaluation method for the reliability of a Markov model switched reluctance motor system. The method comprises: solving a probability matrix P′T(t) of a switched reluctance motor system being in any survival state at any time t via a state conversion diagram of the switched reluctance motor system; calculating the sum of various elements of the probability matrix P′T(t) of the survival state, so that a reliability function R(t) is obtained; and thus calculating the average working time of the switched reluctance motor system before failure, thereby realizing the quantitative evaluation of the switched reluctance motor system and satisfying the requirements for the reliability analysis of a switched reluctance motor drive system. This disclosure has a good engineering application value.