Tooth surface pitting is one of the main faults of helical gear. A large number of experimental data are needed for the diagnosis of pitting corrosion faults on different tooth surfaces and different degrees. Therefore, a method combining simulation and experiment is proposed to identify the degree and quantity of pitting corrosion on helical gear tooth surface. The envelope spectrum analysis of dynamic response signals is carried out, and the characteristic response laws of different pitting fault types in time domain and frequency domain are obtained. Through the characteristic analysis of the vibration signal of gear contact fatigue experiment, the accuracy of the dynamic model of helical gear pitting corrosion fault is verified. The results show that the gear pitting fault dynamic model can diagnose and identify the pitting corrosion degree of single tooth surface and multi tooth surface, and provide a diagnostic basis for transmission gear fault diagnosis and health prediction.