Traditional credit risk earlywarning model based on fuzzy assessment technique has larger functional limitations under the present credit mutation status. This paper empirically analysizes the commercial bank credit risk earlywarning process based on preference entropyweight and matterelement extension model by choosing the real data from different periods of some sample enterprises in Suzhou City and produces empirical results. It is believed that credit mutation makes earlywarning quality of sample enterprise decline slightly, but it never causes earlywarning grade to present a “jump” mutation in a short term, and earlywarning grade of sample enterprise always maintains a mild state, which is enough to show that preference entropyweight and matterelement extension model may well solve the problem concerning commercial bank credit risk earlywarning under the credit mutation.