Regulation (EU) No 347/2013 (the Regulation) sets out the guidelines for the identification, analysis and selection of socio-economic beneficiary energy infrastructure projects called Projects of Common Interest (PCI). To become a PCI and benefit from the Regulation and the Connecting Europe Facility, candidates to the PCI status must provide a cost-benefit analysis in compliance with methodologies developed by the European Networks of the Transmission System Operators for electricity (ENTSO-E) and for gas (ENTSOG). Those Energy System-Wide Cost-Benefit Analysis (ESW-CBA) methodologies are themselves framed by the Regulation, which specify the objectives to achieve and level of detail that the methodologies should incorporate. The European Commission launched this study to review the first version of the ESW-CBA methodology for gas projects. The methodology was published by ENTSOG in 2015 and was used for the adoption of the Ten-Year Network Development Plan 2017 as well as the adoption of the first and second PCI Union lists. The study concludes that the methodology has significant potential for improvement at both regulatory and economic levels. The methodology is too simplistic in its modelling and monetisation assumptions and lacks sufficient level of transparency. This limits its ability to provide sound and unbiased outputs which is needed for the defined decision processes, such as PCI selection, CBCA decisions and evaluation of co-financing requests. The study has identified a series of necessary improvements to cope with the methodological shortfalls. They were designed in collaboration with the key stakeholders and substantiated through a public survey. Special attention was given to the relevance of proposed improvements from an economic viewpoint and to their ability to serve decision-makers. An according update to the methodology will ensure more accurate, reliable and useful output that can be used more directly by decision-makers, while striking a practical balance between the methodology’s accuracy and the cost of its implementation.