The Roads Economic Decision Model is an assessment tool created by the World Bank for the analysis of investments on roads with low levels of traffic in developing countries. Of the variables used in this model, roughness is one showing the highest incidence in the determination of a project's net present value (NPV). The difficulty in defining roughness, a qualitative attribute of this type of road, in a precise figure may lead to biased NPV estimates. This paper introduces the application of a fuzzy roughness index based on simple fuzzy math, which enables the translation of a qualitative, firsthand appraisal of a road's roughness into a quantitative expression—a triangular fuzzy number—and thus increases the amount of information in the model. As a result, the estimation of NPV is also interpreted in terms of fuzzy math. Investment alternatives for a road may be then compared on the basis of the representation of each fuzzy NPV obtained for the investment alternatives. An analysis conducted on the basis of fuzzy roughness and fuzzy NPVs can lead to an alternative that would have been discarded with a standard approach.