Abstract: This paper presents a terrain-aided navigation system which can correct the errors of inertial or other guidance systems and improve the navigation accuracy.This system is composed of a Kalmaa filter,a radar altimeter and a digitized terrain map.Based on the stored terrain data and the state predicted currently by the navigation system,it is possible to calculate the predicted clearance and slope of terrain in real-time.The predicted clearance is compared with the actual clearance measured by radar altimeter in the Kalman filter and the error estimates of the navigation system are calculated according to the terrain slope.Then,the error estimates are fed back to update the navigation system.The process described above is iterated so that the terrain-aided navigation system is updated almost con-tinuosly.The stochatic linearization technique is utilized to obtain the linearized measurement equation.The terrain-aided system can reach a high accuracy for vehicle location.The final estimated errors of position are approximately as follows:here,σn2 is the variance of measurement noise,σsx2 and σsy2 are the variances of terrain slopes in x and y direction respectively,N is total number of updates.The results of computer simulation indicate that the location accuracy of the terrain-aided system can reach magnitude of about ten meters for a typical hilly region.