Abstract: Financial indicators and corporate governance variables were sieved separately to get representative variables via factor analysis, mean value test and correlation analysis. Then, an empirical analysis was done by support vector machine (SVM). The results showed that the SVM model is superior in predicting the financial bankruptcy risk to other methods. Comparing the SVM model with the model based on financial indicators, it is found that the model introducing corporate governance variables in it is more predictable, where the variables include the proportions of circulating shares, shares held by the biggest shareholders and share ownership concentration. This method is worthy of practical applications to a certain extent.