For the problem that the traditional angle sensor’s complex mechanical structure and prone to failure and the gyro bias cause the error accumulates over time in automatic driving system, a wheel turning angle measurement system based on double GNSS antennas and single gyro was proposed. The system’s sensors mainly included two GNSS antennas and a MEMS gyro. The double GNSS antennas were mounted on both side of the vehicle and provided the speed, attitude angle, latitude and longitude of the vehicle. The single MEMS gyro was mounted on the wheel and the angular rate was measured. An algorithm used the above data was designed to integrate the angular rate to obtain the steering angle. In order to solve the problem that the error accumulates over time, a Kalman filter based on the vehicle dynamics model was designed to calibrate the error of integration of the gyro data. At the same time, the lever-arm compensation algorithm was used to solve the speed error caused by lever-arm. The straight line experiment was carried out to verify the effectiveness of the system and the curve line experiment was carried out to verify the effectiveness of the lever-arm compensation algorithm. Compared the steering angle of the proposed system with the Hall effect angular sensor, the average error of the straight line experiment was -0.064°and the error variance was 0.309°and the cure line experiment’s mean error was 0.299°and the error variance was 1.009°. The result of the experiments showed that this system could replace the traditional angle sensor and it was easy to install and overhaul.