Power grid operation and maintenance costs form important part of power transmission and distribution costs.Accurate and rational cost data are needed in order to calculate the price scientifically.However，there are significant differences from province to province in the operation and maintenance costs.It is difficult to determine their rationality by making a unified standard or classification standard.This paper studies the rationality evaluation based on the theory of statistical regression analysis.The main influence factors are firstly filtered by analyzing the influence factors.Secondly，the principal components are extracted from the main factors to eliminate the multiple correlations.Then statistical regularity of power grid operation and maintenance costs is obtained after regression analysis.Finally，the Chauvenel criteria are used to evaluate the corresponding rationality.Through the case studies with the actual data of27provincial grid enterprises，the results show that the method is effective and feasible.