[en] As part of the 'Probabilistic Safety Assessment of safety grade digital systems used in Nuclear Power plants' research, measures and methodologies applicable to quantitative reliability assessment of safety critical software were surveyed. Among the techniques proposed in the literature we selected those which are in use currently and investigated their limitations in quantitative reliability assessment. One promising methodology from the survey is Bayesian Belief Nets (BBN) which has a formalism and can combine various disparate evidence relevant to reliability into final decision under uncertainty. Thus we analyzed BBN and its application cases in digital systems assessment area and finally studied the possibility of its application to the quantitative reliability assessment of safety critical software