Falling weight deflectometer (FWD) test has been used widely for pavement condition assessment, and different back-analysis techniques and commercial software have been developed to predict the layers’ moduli. Despite the recent use of FWD tests for railway substructure condition assessment, there have been no reported instances of commercial back-analysis software specifically designed for this purpose and only a limited number of research-based techniques exist. Among the available back-analysis packages, BAKFAA stands out as a notable option, which was originally developed for pavement applications and asserts its adaptability for railway applications. This paper aims to investigate the suitability and accuracy of utilizing BAKFAA for railway applications, assess its performance, and determine whether it can effectively meet the requirements of railway analysis. In this study, the primary developed and validated finite element (FE) model of a railway track section by the authors is employed to generate extensive virtual experimental FWD testing data. In this regard, various layers’ moduli for each layer are considered over a predefined range to represent the real field conditions. Then the back-analysed values through BAKFAA are compared against the target values (Cone Penetration Test (CPT) data) to check the accuracy of the BAKFAA predictions. The findings of the study indicate that BAKFAA tends to overestimate the moduli of substructure layers. To address this issue and align BAKFAA with the virtual experimental data, the study proposes a correlation model. This model can enhance the applicability of BAKFAA for railway applications by making necessary adjustments based on the virtual experimental data.