Abstract: It is unable to analyze the reliability of the composite transmission and substation system which incorporates the equivalent multi-state substation components by using traditional cross-entropy sequential Monte Carlo (CESMC) method. Therefore, this paper first analyzes the number of equivalent states of the substation, and derives the optimal reliability parameter expression and parameter optimization iterative formula for the equivalent multi-state components of the substation based on Markov and cross-entropy importance sampling theory. A modified CESMC reliability assessment method suitable for the transmission and transformation joint system is proposed, which expands the scope of application of the traditional CESMC reliability assessment method and solves the problem that sequential cross-entropy can not be applied to multi-state system reliability assessment. Combined the improved CESMC method, the reliability assessment of the power transmission system containing equivalent components of the substation is accelerated, and a reliability assessment model for the transmission and transformation joint system is further constructed. Finally, as a case for testing the proposed model, the reliability of IEEE-RTS79 system is analyzed. The simulation results show the availability and the advantages of the proposed method when assessing the reliability of the composite system. Meanwhile, the influence of the substations on the reliability of the transmission system can also be found according to the results.