In the future smart grids, Plug-in Hybrid Electric Vehicles (PHEVs) are seen as an important means of transportation to reduce greenhouse gas emissions. One of the main issues regarding to this sort of vehicles is managing their charging time to prevent high peak loads over time. Deploying advanced metering and automatic chargers can be a practical way not only for the vehicle owners to manage their energy consumption, but also for the utilities to manage the electricity load during the day by shifting the charging loads to the off-peak periods. Additionally, an efficient charging schedule can reduce the users' electricity bill cost. In this paper we propose a new practical demand response (DR) program for PHEVs charging scheduling based on game theoretic approach, aiming at optimizing customers charging cost. In the proposed method, a stochastic model is given for starting time of charging, which makes the method a practical tool for simulating the vehicle owners charging behavior effectively.