This paper proposes a 2-step Kalman filter model for land vehicle navigation using civilian-band GPS measurements. The velocity bias caused by the Earth#s rotation would be removed completely when applying this model. Because the linearization of velocity equations in this model is not necessary, the error is significantly reduced. The experiment reveals that estimated position error with stationary data is about 5m during a 15-20 minute interval. The other benefit of this model is that it can be feasibly applied as a GPS receiver module thanks to the small sizes of the necessary manipulating matrices.