Abstract: Probabilistic safety analysis（PSA） is one of the key methods to ensure the safety of nuclear power plant operations. The current practices of PSA mainly rely on limited data and expert experience to conduct static risk quantification, and hence cannot accurately reflect the actual risk of nuclear power plant operations. A large amount of data has been collected and stored in real-time through the nuclear power plant operations, which can be used to track the system safety status, referred to as, safety big data. In order to improve nuclear power plant safety, it is of great significance and urgent need to research and develop the technology and method for real-time PSA by integrating the safety big data into the current PSA methods. Therefore, this paper briefly introduces the status of safety big data in the safety assurance of nuclearpower plants and focuses on the research status and prospects of real-time PSA of nuclear power plants driven by safety big data. The results show that safety big data would contribute to a comprehensive analysis of the operation safety of nuclear power plants in real-time; it would be the important direction of future research to systematically integrate the safety information extracted from safety big data and the current PSA methods to achieve accurate and trustworthy real-time safety analysis of nuclear power plants.