Urban infrastructures are characterized by considerably high investment costs, which make investments almost irreversible. Once the construction is completed, it is difficult and expensive to modify the systems. Therefore a detailed cost analysis should be carried out in advance, to prevent unprecedented future charges and provide reliable infrastructure services. This study aims to reveal the economic characteristics of urban energy (electricity and natural gas) distribution systems' investment costs using (1) capital investment costs at the local level; (2) outputs such as sectoral numbers of customers and energy sales, and input prices; (3) company-specific characteristics of energy distribution firms; and (4) socio-economic and site-specific urban and geographic variables. However, the fourth group variables have often been neglected in the literature, although the impacts of such variables are expected to be highly relevant. In this study, an econometric approach has been utilized for cost analysis. Regression models tested both log-log and the Box-Cox forms, while the Box-Cox regression model, as a flexible form allowing for the endogenous determination of the parameters, gave better results. Regression estimates show that besides the output and input variables, and company-specific variables, socio-economic and site-specific urban and geographic variables have statistically significant effects on electricity and natural gas distribution capital investment costs. Moreover the results provide evidence for economies of scale and density in these distribution systems under various configurations of market size and local conditions.