The invention belongs to the technical field of road maintenance and detection, and particular relates to a method for testing pavement smoothness based on a smart phone. Built-in equipment of the smart phone comprises a triaxial acceleration sensor, a gyroscope, a magnetic sensor, an azimuth sensor, GPS sensing equipment and data acquisition equipment. The method specifically comprises the stepsof step (1) acceleration data acquisition, step (2) acceleration data calibration, step (3) acceleration data smoothing and denoising, step (4) space-time coordinate conversion algorithm and (5) BP neural network calculation model establishing, wherein a BP neural network model uses a three-layer neural network, the input vectors of sample data are acceleration and vehicle speed, and the output vectors are international flatness index IRI. According to the method, a BP neural network method is utilized to establish regression prediction models of acceleration values and IRI under different detection speeds, and the models are optimized according to different applicable conditions. The model is high in prediction accuracy and good in stability, and can accurately reflect the relationship between the acceleration values and the IRI.