Several studies have found that cost-benefit analysis (CBA) of transport projects has a limited impact on project selection, even if decision-makers claim it has. One potential reason for this could be that CBA is conducted too late, making it politically difficult to filter out projects with a low NBCR (net benefit-cost ratio). A possible remedy for this problem is to identify general project-specific characteristics that affect social costs and benefits at early stages—during the front end of the project—when there is greater flexibility to take different courses of action and the economic and political costs of making changes are relatively low. However, to our knowledge, no previous research has identified such project-specific characteristics, despite the growing interest in this idea within project management literature. One possible reason for this gap in the literature could be a lack of data—few researchers have access to detailed data on a large number of transport project CBAs. This paper contributes to the literature by empirically identifying which project-specific characteristics, available at early stages, are associated with high or low NBCR in a project. It uses data from over one thousand road projects from four national transport investment plans in Sweden and Norway.