A binary decision diagram (BDD) is a data structure that is used to represent a Boolean function. Converting fault tree into BDD can effectively simplify counting processes and improve the accuracy and effectiveness of the results. However, due to various types of uncertainties in reliability data, we cannot obtain precise failure probabilities. In order to accurately quantify the certainties and obtain much more reliable results, we use BDD method based on fuzzy set theory for reliability quantitative analysis. In this regard, we take W-axis feeding system of heavy-duty computer numerical control (CNC) machine as a project example and adopt fuzzy BDD quantitative analysis method to analyze its reliability. The analysis results (aided by computer calculation) illustrate the effectiveness of the method proposed in this paper. © 2016, Shanghai Jiaotong University and Springer-Verlag Berlin Heidelberg.