Risk analysis is regarded as one of the most critical research topics in maritime industry due to the nature of work. Validation of risk analysis is a significant challenge for safety researchers. This paper aims to address the gap by utilising the power of Bayesian Network method under fuzzy logic environment. While Bayesian Network provides a compact representation of a joint probability distribution, fuzzy logic tackles with vagueness and ambiguity in PRA (Probabilistic Risk Assessment). The validation and sensitivity analyses are performed under various conditions to improve the accuracy of findings. Three different axioms are tested to validate the results. The proposed method is applied for the risk modelling of ship collision in narrow waters. Besides its theoretical background, the paper is expected to help safety researchers validate their results. The maritime safety professionals (safety inspectors, DPAs, superintendents and ship owners) can also benefit from the paper to identify and minimise the potential risks for ship collision in narrow waters.