This paper proposes a position accuracy improvement algorithm for a low-cost global positioning system (GPS) receiver of a car navigation system considering a distance boundary. The proposed method enhances the position accuracy by matching the trajectory of a vehicle acquired by a GPS receiver with prebuilt map information. After acquiring the trajectory of a vehicle, the iterative closest point method is employed to compute the rotational and the translational errors among the trajectory of a vehicle and map information. These computed errors are used to correct the trajectory of a vehicle. Unlike previous studies, in the proposed method, some of map information over a certain distance away from the trajectory is ignored so that the computational efficiency can be guaranteed. Finally, the superiority of our method is proved throughout real experimental results.