This study proposes an integration strategy regarding the efficient prediction of hotel bankruptcy by combining data mining techniques. In particular, by combining support vector machine(SVM) and neural network(NN), SVM based NN hybrid model for hotel bankruptcy prediction is newly introduced in this study. In the experiments on Korea deluxe hotel data, SVM-NN hybrid model achieves a performance accuracy of 96.34%, which is better than that of stand-alone classifiers. The hybrid model performs better in the grey area where some bankrupt hotels appear to be less financially distressed. The results suggest that debt-burdened hotels with low profit margin and ordinary income margin as well as lower growth in asset are more likely to be candidates of bankruptcy. Accurate bankruptcy prediction usually brings into many benefits such as risk reduction in investment return, better monitoring, and an increase in profit. Limitations of the study and avenue for future research directions are also discussed at the end.