本文旨在利用公司會計資訊為基礎的信用風險衡量技術，建構企業（尤其是非家族業）信用風險模型（Credit Risk Models），找出影響公司績效之顯著變數，儘可能在公司尚未發生信用風險前，預測公司可能發生信用風險之機率，降低財務風險損失，達事前預警與風險管理之效。實證分析上，利用會計資訊變數做為解釋變數，進行K-S檢定、M-U檢定與建立Logistic信用風險模型。實證發現，會計資訊變數模型一之風險敏感度高於複迴歸模型二；會計資訊變數影響公司績效之主要指標有財務結構、償債能力、經營能力、獲利能力、現金流量指標；增加自有資金比率、存貨週轉率、股東權益過轉率、營業利益率、業外收支率、每股現金流量，降低稅後淨利率，將降低公司發生信用風險機率。研究建議，提高營業利益率使其大於業外收支率、降低內部保留比率（增加固定資產）、調整融資決策（增加流動負債）、提高稅前淨利率，可增強公司獲利能力；控制固定資產週轉率和現金流量比率，可降低信用風險發生機率。實證價值與管理涵義，銀行提高企業授信品質，可降低信用風險發生之可能性，降低風險性資產比重，進而提高資本適足率。

This paper primarily used accounting data to construct a logistic credit risk models on non-family controlled companies. The credit risk model predicted in advance the probability of financial risk attacking non-family companies. We picked out the significant variables as key indicators of corporate performance. In the empirical analysis, we built up a logistic credit risk model using accounting data, K-S test and M-U test. Finding, the risk sensitivity were the accounting data modelIsuperior the multiple regression model II. The key performance indicators were the financial structure, the solvency, the operating performance, the profitability, and the cash flow. The companies increasing the own funds ratio, the inventory turnover, the equity turnover, the operating margin, the ratio of non-operating revenue and expenses, the cash flow per share, and reducing the after-tax margin could reduce credit risk. Empirical results suggested that enhancing the operating margin more than the ratio of non-operating revenue and expenses, reducing the retention ratio, adjusting financial decision, and enhancing the ROS, all could enhance the profitability. Controlling the fixed asset turnover and the cash flow ratio could reduce credit risk. Implications, a major part of the business of financial institutions must follow to minimize credit risk and make successful loans could reduce risky asset and enhance BIS ratio.