This paper describes a design of tracking algorithm for a land vehicle using stand-alone GPS. The algorithm comprised three major parts - a vehicle motion change detector, acceleration estimator and finally a Kalman filter with feedback. It contains two new techniques, acceleration estimation using a sliding mode and a vehicle motion change detection algorithm, which are developed to improve navigation accuracy. The algorithm provides more accurate position information than the conventional iterated least-squares solution, and it can easily be applied to a real-time route guidance system. Real experimental results are presented to demonstrate the advantages of the proposed algorithm. (C) 2000 Published by Elsevier Science Ltd.