In order to solve the problem that the body of potato combine harvester cannot self-adaptive balance with the change of terrain, which leads to low operation safety, big harvest damage and poor harvest quality, a body leveling system of potato combine harvester was designed.In this system, the inclination sensor of the first-order inertial filter was used to monitor the lateral inclination of potato combine harvester body, and the interference and jitter were effectively suppressed. The mathematical model of the system was established through the dynamic analysis of the body leveling system mechanism. The fuzzy PID algorithm based on the first-order inertia filter was used to control the electromagnetic proportional valve to drive the lifting hydraulic cylinder to realize the automatic leveling of the potato combine harvester body.The simulation results of the body leveling system showed that fuzzy PID algorithm had better control performance than traditional PID algorithm, which can effectively shorten the system regulation time by 51.77%, the rise time by 53.57% and reduce the maximum overshoot by 6.25%. The static and dynamic experiment results of the whole control system showed that in the range of -10°~10° slope, the automatic leveling time of the system was less than 4s, and the maximum leveling error was less than 1°. Under the condition of 10° incline, the auto leveling time using the fuzzy PID control algorithm was reduced by about 50%. The static test results were consistent with the simulation results. When driving at 3.6km/h speed on the road with great fluctuation, the error of vehicle body tilt angle was controlled within ±3°. Therefore,the automatic leveling control of potato combine harvester was well realized, which met the actual operation requirements.