China Railway Express (CRE) is one of the most important constituent parts of the Belt and Road Initiative. Freight demand analysis is fundamental as a basis for the operational strategy of CRE and the investment policy along the CRE-routes. Most of the existing relevant literature has focused on the organization of the train operations of CRE, with little research related to demand analysis. This paper contributes to filling this gap by estimating customers’ demand preferences for rail freight service attributes, by using a novel multi-criteria decision-analysis (MCDA) method namely Best-Worst Method (BWM). To this end, a BWM survey was conducted in China to capture customers’ preferences for the main attributes that define the transport service offered by CRE. Two variants of BWM, the linear and the Bayesian, are employed for the analysis. Reliability is suggested as the most important attribute for CRE to focus on, to gain customers. We also conduct a cluster analysis based on the results, which helps the CRE operator to identify homogeneous customer segments, and to optimize the use of CRE's resources with a differentiated pricing strategy.