Due to the effect of shielding from buildings and multipath interference,GPS/INS integrated navigation system is unable to meet the demand of practical application in urban environment.To solve this problem,a GPS/INS integrated urban navigation algorithm based on vehicle motion detection is researched.The algorithm provides vehicle estimated information such as relative position and attitude angle by the integrated Kalman filter.At the same time,a motion detection module is used to estimate its motion performance.This information can help the system locating and improve location precision when the signal of GPS is lost.The car test with low cost sensors shows that the system has high accuracy and can restrain the error cumulate when GPS without signal.So the system can satisfy the needs of vehicles navigation in urban environments.