Credit risk assessment is the groundwork of the individual housing loans. However, what banks really care about is how to make the most money with least risk. The most difficult problem of the credit risk management for the banks is that they both have to encourage loans, but also they need to avoid bad debt. This paper proposes a new risk assessment system of the individual housing loans by integrating traditional Analytic Hierarchy Process and Grey System Theories into Data Mining and applies the new hybrid method to assess the individual credit risk. It is proved that this new credit risk assessment method is easy to calculate, and provides an effective foundation for credit risk assessment.