Abstract: Aiming at the current problems of large transmission error, high contact temperature and high contact Hertzian stress of the spiral bevel gears, an optimization method based on optimal parametric modeling and non-dominated sorting genetic algorithm is proposed. Firstly, based on Kiss soft software, the spiral bevel gears are parametrically modeled to clearly describe their parameter variation characteristics. Secondly, the spiral angle, pressure angle and tooth width are taken as the optimization variables, and the transmission error and longitudinal overlap are taken as the objective functions, constructing a multi-objective parameter optimization design for the spiral bevel gears based on the second-generation non-dominated sorting genetic algorithm (NSGA-II) . Simulation results show that the longitudinal overlap of the optimized gear pair is increased to 21.92%, while the transmission error, contact temperature, and contact Hertzian stress are reduced by 46.90%, 23.69%, and 32.19%, respectively. The proposed optimized design significantly improves the smoothness of the gear pair transmission, increases the gear life, load carrying capacity, transmission efficiency and reliability, and effectively reduces the vibration and noise of the gears.