The article systematically analyzed the actuality and various quantitative analyses methods of Chinese listed corporations’ financial distress. At present, domestic and overseas scholars have researched the reason and discussed the reason of this problem appeared and the method how to resolve. At the same time, the development of statistics, mathmatics and Artificial Intelligence has made it possible that apply the compositive data mining theory to this field. This article generally analyzed the existing financial crisis forwarning models, and established two forewarning models using Enterprise Miner module in SAS.The models in this article possessed includes Logistic models and Neural Network model possessed many advantages and obtained perfect forewarning effects.Part of the introduction of Chapter one of this artical explained the concept of financial distress, and defined the company that is at financial crisis as "ST" company. At last, this part discussed the studying purpose, meaning of this subject. The research emphases and research path were also explained.Chapter two studied mainly the relative theory of financial distress. First, this part pointed out that applying the quantitative analysis’s methods to financial analysis fields has become an important research method, but the limitation also existed. Then, studying actuality both at home and abroad were reviewed and analyed. This part also compared the existing models based on traditional statistic methods and discussed the existent limitation. Based on above, the idea was put forward that with the development of computer science and artificial intelligence, more compositive data mining theories and methods would be increasingly applied to the research on financial distress.Chapter three explained the basic theory of data mining, included conception, categories and major purpose. The process model of data mining was the keystone of the part. The SEMMA model was discussed which provided a technical implement measure for the process of data mining.Chapter four discussed the major methods of data mining. The traditional statistic methods such as Logistic regression analysis and the rising Artifical Intelligence methods such BP Neural Network were introduced. How to apply these theories to financial distress forewarning purpose was also introduced.Chapter five was an empirical study. Based on the SEMMA model, the chaptermade use of financial data of Chinese listed corporations and Enterprise minermodule in SAS to realized those arithmetics.Chapter six reviewed the artical’s major viewpoint, achievement and limitation. The expectation for the future research was also put forward.