Abstract: Comparing the Bayesian theory method and the traditional means, the different results validate the rationality and practicability of applying of the bayesian means to the item repair. Based on analysis on the process of replace repair of gearbox, concentrative item repair is confirmed as the key process. According to the fact of replace repair, i.e., the mean value and standard deviation is changeable with the environment, the Bayesian theory is adopted, in which the mean value and standard deviation is deemed as stochastic. The Bayesian estimate of posterior distribution is obtained from the history data and prior distribution, furthermore, the key process capability is analyzed by using SPC technology with the 97.5% one-sided confidence lower limit of process capacity index given with comparison with classical theory. The comparing result indicates that the applying of Bayesian means to the item repair is more rational and practical than the traditional method.