The paper discusses bankruptcy prediction model in the UK during the two last decades. My study is provided to support that the Original Altman's Z-score (1968) might not valid to predict bankruptcy since the business environment changed a lot. However, there are many firms go to bankrupt recently and there is a need to study and improve the bankruptcy predictive ability. And the result shows that Altman's Z-score has little predictive ability in bankruptcy prediction. Then, I use the recent data to renew the Z-score model by changing the coefficient of original Z-score. After compared to the original Altman's Z-score model, I found that the renewed Z-score model has been improve to a reasonable accuracy rate. In addition, I found that the variable (Sales/ total assets) has little contribution to distinguishing the bankrupt and non-bankrupt firms.