Abstract: Abstract: The reliability of a kind of drive axle housing was studied mainly in this study. The three-dimensional parametric model of the drive axle housing was established by using parametric design language of ANSYS (APDL), and then the strength and stiffness reliability were built on the basis of the interference principle of reliability. The reliability analysis of drive axle housing was discussed under the following four typical working conditions: maximum vertical force, maximum tractive force, maximum braking force and maximum lateral force. The strength and stiffness reliability under the conditions of different working were analyzed respectively. Considering the randomness and uncertainty of the design parameters, the geometry size, load and material strength were set into normal distribution (GAUS in ANSYS) and uniform distribution (UNIF in ANSYS). The Monte-Carlo sample method and stochastic finite element method were adopted to carry out the reliability analysis. A simulation cycle was equivalent to a sampling test, and the longer the simulation cycle lasted, the larger the sampling number was, and the better accurate the simulation result was. With the simulation cycle time of the four conditions of 488, 649, 928 and 488 respectively, the mean and standard deviation of equivalent stress sampling and deformation sampling were stabilized. The results mainly included the static analysis, the reliability analysis and the sensitivity analysis. Firstly, the equivalent stress contour and deformation contour of the drive axle housing under these four conditions were shown. Secondly, the strength reliability and stiffness reliability was also obtained. The strength reliability of all conditions and the stiffness reliability of maximum vertical force condition were more than 90%, and the stiffness reliability of the other three conditions was 100%. The results also showed that the strength reliability of the maximum lateral force condition and the stiffness reliability of maximum vertical force condition were the worst. The maximum vertical force condition and the maximum lateral force condition should be given priority consideration in the design process. Finally, through the sensitivity analysis results, the influence rules of design parameters on the reliability of the drive axle housing were obtained. The most important parameters included dynamic load coefficient K1, maximum vertical force of full load in the static state F, maximum tractive force P, material strength S and geometry size L, L1 and D4. When the parameters of K1, F and P became larger, the reliability of the drive axle housing would turn worse. On the contrary, the reliability became higher with the increasing of S, L1 and D4. Additionally, the effect of the parameter L (half of wheel track) on the reliability of the drive axle housing was more complicated. The reliability became better with the increasing L under the maximum vertical force condition and became worse under the other three working conditions. Therefore, the change of L should be strictly controlled in the design process. The research method and the analysis results can provide a theoretical guidance for the reliability design of key parts such as drive axle housing of the agricultural machinery chassis.