The process to measure the slop and cross-fall of highway with traditional manual methods has disadvantages of speed, low accuracy, and low efficiency, which needs imperative traffic control and shows several shortcomings in real time, flexibility and security. We design a laser detection arrays composed of four laser range finders (LRF), and an inertial measurement unit (IMU) is combined with global positioning system (GPS) to form a measurement plane, which realizes the measurement of slop and cross-fall of highway. Then a calibration method based on relative movement is proposed, with which we can obtain the deviation angle of laser detection arrays and correct measurement data for calibration, then we can improve the measurement accuracy. The actual road test shows that with the proposed method the accuracy is improved about 5% after calibration. The repeatability of the slope and cross-fall experimental data is more than 95%, and the correlation reaches 99.9%, which can meet the requirements of the standard of field test methods of subgrade and pavement for highway engineering.