Abstract: Expert systems are rapidly growing in diversity of application and usage worldwide. To enhance the security of nuclear power plant systems and help the operator for analysis and decision-making in the event of a failure, a nuclear power plant fault diagnosis expert system based on dynamic uncertain causality graph (DUCG) was built. Design and function implementation of the system knowledge base, inference engine and graphical user interface were introduced in detail in this paper. This system is proved to be capable of accurate fault diagnosis based on the verification results and provide support for the operator decisions.