Refuelling facilities are critical infrastructure for road freight transport networks, since every vehicle must refuel in order to move freight.Recent advances in alternative fuel (alt-fuel) vehicles have made the location of refuelling facilities even more critical, since the driving range of alt-fuel vehicles is less than for traditional fuel-powered vehicles (Deb et al., 2018).Transport companies have decided to establish their own refuelling facilities in order to guarantee the feasibility of their vehicles' movements, as well as to optimise the refuelling process.This article presents a methodology to assist transport companies to locate their own refuelling facilities.It first proposes a methodology to convert GPS data into trips to understand where the vehicles are moving, then uses this information as an input for a customised Mixed Integer Programming (MIP) model that suggests potential locations for refuelling facilities.The methodology is applied to a case study in Australia (the Melbourne-Sydney corridor).The transport company that is the subject of this case study provided all the required data to run the model 1 .Their motivation to do the analysis in this corridor was the high amount of fuel refuelled at retailers (refuelling facilities of third parties), which is in the order of tens of millions of litres of fuel per year.They decided to investigate the movements of the fleet to evaluate the business viability to locate their own-refuelling facilities in the corridor.The main advantage for the transport company to have their own-refuelling facilities was to enable the purchase of fuel at wholesale prices directly from fuel suppliers, thus improving the efficiency of refuelling operations, since strategically located refuelling facilities could mean that the vehicles spend less time in the refuelling operation.For this particular case study, the model suggests that only by improving the refuelling decision taken by the fleet (where and how much to refuel) 2.3% of savings (representing hundreds of thousands of dollars) could be obtained out of the total refuelling cost of the company in the analysed corridor.After doing a sensitivity analysis on the fuel consumed in the corridor, it is suggested that only one own-refuelling facility should be located, since the location of two refuelling facilities would not be an optimal solution if the fuel consumed in the corridor is reduced by more than 20%.The MIP presented in this article is a contribution to the flow r efuelling l ocation l iterature, s ince i t adds differential prices of the fuel in each refuelling facility as a driver to locate refuelling facilities.Another contribution is the novel procedure presented in this paper to process and convert large amounts of GPS data into trips.