Since year 1997 up to 1999, 64 banks in Indonesia have been liquidated/closed by Central Bank and numbers of bank have followed banking re-capitalization program or under control of Indonesian Banking Restructuring Agency/OKRA. It brings Indonesia into a big economic problem because Bank's role in Indonesian economy was very important. Indonesia becomes a country which has a big burden and very hard to recover from the crisis. That is the reason that the author intends to explore more precisely the condition of Indonesian Banking.This research is designed to make a model that can be used to predict Bank Failure with case study in Indonesian Banking 1997 - 1999. This research purpose is to differentiate good bank and failure bank, so the result of this research will contribute to banking knowledge in Indonesia and will help the Indonesian to select a good bank.This research uses Multiple Logistic Regression Model as a methodology with more than 2 (two) independent variables. The model had been chosen since the type of dependent variable is binary (Good bank or Failure bank) and allows us to use 'dummy data' that can not be possible in conventional model.The numbers of data sample in this research were 144 banks : 81 failure banks and 63 good bank. Data used in the model was the last published financial statement before the bank failure. The dependent variable is failure/good bank and there are 16 independent variables related to Capital, Asset Quality, Earning/ Profitability, Liquidity and Efficiency.The model was tested by determination coefficient Cox & Snell R Squared and Nagelkerke R Squared. The test shows that the model is significant (0.6 & 0.8). it means that independent variable used in the model has significant correlation to dependent variable. Kolmogorov Smimov Test shows that the model be able to differentiate failure bank and good bank. Hosmer & Lemeshow Test proves that the prediction of the model is fit with the actual data. The above tests summarize that the model can be used to predict and differentiate between failure bank and good bank.