Abstract: Digital instrumentation and control system (DICS) has been widely applied to safety/non-safety systems in NPP. As the central nervous system, its reliability is crucial to the operation and maintenances in the process of operation. Based on the structural and control logics of the digital steam generator water level control system (SGWLCS), the Boolean logic driven Markov process (BDMP) formalism was recommended and applied to the assessment of the system. With the increase of simulation steps, the cumulative distribution probability (CDP) of repairable and non-repairable SGWLCS was calculated. The BDMP model was constructed by KB3, and the productivity analysis was performed by the Monte Carlo simulation software YAMS. The tendency of the system CDP used to derive the failure probability with the performance period was assessed in quantitative way. Through the sensitivity of each component of the system, the importance to the whole system of each component can be intuitively achieved.