The assessment of transportation social sustainability appeals to many scholars as an essential issue but imposes a risk of selecting inadequate factors and methods to measure social phenomena. In this study, based on an extensive review, we identify the measurable key social indicators (instead of the economic or environmental factors commonly used) and propose a comprehensive evaluation framework for relative performance analysis of social sustainability in the regional context. This work is the first attempt to assess regional social sustainability of transport explicitly utilizing a multi-output performance measure. We use Shannon entropy to combine the results of selected data envelopment analysis (DEA) models into a unified social sustainability performance score. The method is applied to regional road transport in the European Union (EU) member states for the period 2004–2017. The empirical section explores individual profiles of EU countries, compares the states grouped into two clusters of old (EU-15) and new (EU-13) EU members, and examines their social sustainability performance over time. The analysis suggests that considering only social factors in the measurement eliminates the bias resulting from the inclusion of economic factors. As a result, our method prevents inaccurate inflation of the scores of more economically advanced countries. The findings also draw attention to the car dependence problem associated with high motorization rates in affluent EU-15 states. The study confirms the capacity of the proposed DEA-based framework to serve as an adequate tool for measuring the social sustainability of transport, which can support policymakers by providing useful benchmarks employing social factors.