[en] Despite the efforts to avoid undesirable risks, or at least to bring them under control in the world, new risks that are highly difficult to manage continue to emerge from the use of new technologies, such as the use of digital instrumentation and control (I and C) components in nuclear power plant. Whenever new risk issues came out by now, we have endeavored to find the most effective ways to reduce risks, or to allocate limited resources to do this. One of the major challenges is the reliability analysis of safety-critical software associated with digital safety systems. Though many activities such as testing, verification and validation (V and V) techniques have been carried out in the design stage of software, however, the process of quantitatively evaluating the reliability of safety-critical software has not yet been developed because of the irrelevance of the conventional software reliability techniques to apply for the digital safety systems. This paper focuses on the applicability of Bayesian Belief Net (BBN) techniques to quantitatively estimate the reliability of safety-critical software adopted in digital safety system. In this paper, a typical BBN model was constructed using the dedication process of the Commercial-Off-The-Shelf (COTS) installed by KAERI. In conclusion, the adoption of BBN technique can facilitate the process of evaluating the safety-critical software reliability in nuclear power plant, as well as provide very useful information (e.g., 'what if' analysis) associated with software reliability in the viewpoint of practicality