The road network design problem is widely studied in the literature, both in urban and rural areas; most papers deal with the solution algorithms and with the mathematical formulation of the problem. The road network design problem is NP-hard and is very difficult to solve in acceptable computing times for real-scale networks. In urban areas, the objective of the design is usually the minimisation in total travel time on the network, while in rural areas it is usually a sum of total travel time and total (i.e. construction and maintenance) road costs. Very few papers consider other terms in the objective function and/or propose a multicriteria approach. In our best knowledge, no papers consider as objective of the design to minimise the fuel consumption, even if it is an actual objective of road network planning and design. In this paper, we compare the results obtained by solving some road network design problems adopting the classic objective functions and adopting as objective the minimisation in fuel consumptions on the network. In order to obtain this result, in the paper a fuel consumption function is proposed, based on CORINAIR models and depending on the vehicle fleet of the study area; this function is used both for estimating the reduction in fuel consumption due to the design obtained with the classic objective functions and for designing the road network using as objective function the total consumption on the network. The results obtained with these approaches are compared in terms of fuel consumed and total travel time spent on the network. The minimisation in fuel consumption is very important since greenhouse gas emissions and air pollutants are directly related to the consumption. Initial tests highlighted that the results obtained with the two different approaches are more significant in rural areas, where higher average flow speeds may produce higher consumption, according to the CORINAIR models.