Abstract: The level of team situation awareness (TSA) is an important reason impacting team reliability. To evaluate TSA level more objectively and quantitatively in digital nuclear power plant (NPP), a TSA causal concept model was constructed by qualitative analysis and expert group discussion, then a TSA reliability evaluation method was developed on the basis of Bayesian network. It not only considered the relative weight of performance shaping factor (PSF), but also developed the methods for determining the conditional probability of intermediate variables and binary variables, respectively, so that the obtained probability data was more objective and reasonable. The specific application of the method was illustrated by case analysis. The results show that the developed method not only could simulate the causal relationship between PSF and TSA reliability, but also could quantitatively evaluate the reliability of TSA under a given accident scenario, and could identify the most important causes triggering TSA errors, which provides theoretical support for the prevention of human error.