Debt rating of small enterprises means a system that evaluates the size of a small enterprise’s credit level and LGD (Loss Given Default) . It concerns the risk management of bank loans and whether small enterprises can get financing. Therefore, the indicators of credit rating must be able to identify the status of non-compliance of small enterprises. Existing credit rating system, aiming to rate scale and sort, does not evaluate the size of a debt LGD, and does not give evidence that its rating system is relevant to identify non-compliance. This article selects a index system according to the ability to identify the default status of enterprises, and establishes a credit rating system of small enterprises. This article has some innovations and features. Firstly, the paper deletesan index whose F-value is small and whose ability to distinguish the status of non-compliance of small enterprises is weak between one pair of highly related indicators which have partial correlation coefficients greater than 0. 7 and reflect the duplication of information. This avoids information redundancy of the rating system after first section and at the same time avoids mistakenly deleting the index that has a big influence on the status of non-compliance. This paper improves the situation that existing credit rating system’s selection is unrelated to the ability to identify non-compliance. Secondly, the paperbuilds the Wald statistic to test for the significant of the Probit regression coefficient, β, through solving the regression coefficient, β, and the standard error, SEβ, of the regression coefficient between the default state variables and evaluation, and deletes the index that has a small influence on the status of non-compliance and whose regression coefficient, β, is not significant. The index can distinguish significantly the status of non-compliance of enterprises after the second selection. Thirdly, the AUC area of the indicator system’s experience curve built by this paper is greater than 0. 9, and this guarantees the index system has a strong identification ability for the status of non-compliance. Fourthly, the results show the weight of the non-financial indicators, for small enterprises having a higher the credit rating and lower LGD, in the credit rating system is 56%. Finally, the empirical analysis for 1 231 small enterprises shows there are 23 indicators, involving the quick ratio, total asset growth, industry sentiment index, which can be used to distinguish the status of non-compliance of small enterprises.