Plug-in hybrid electric vehicles (PHEV) have an EV mode driving range which can cover a portion of customer daily driving. This EV mode range affects the refuel frequency substantially compared with conventional vehicle. For a conventional vehicle, generally the distance between fuel fill up is dependent on tank size and fuel economy, while for PHEVs, distance between fuel fill up is dependent on tank size, fuel economy, and EV driving range. Consequently, the EPA label range does not accurately represent real world driving range between fill-ups for PHEV. Furthermore, for PHEVs, the dependency on EV driving range varies greatly on customer trip lengths. Hence, it becomes critical to use real world customer usage pattern to estimate distance between fill up and size the fuel tank accordingly. This paper describes a methodology to use real world customer data to estimate distance between refueling for PHEVs. The target is to estimate PHEV refuel distance given specific parameters such as EV range, hybrid fuel economy, and tank size. A linear model is developed based on the sensitivity analysis, and a neural network based estimation model is proposed to further capture the non-linearity in the model and improve accuracy.