The policy makers in Indonesia is more Ad-hoc and myopic (the radar far) to make policy. With various models and methods can be made a prediction or forecast about the financial and banking crisis, and even can diagnose the health of can economy. The crisis that can accur in various parts of the world causing expert 1o try to find the cause, a common pattern (stylized facts) that is experts to developed a method to detect the financial crisis, it can even be used to see in the future economy performance. Crisis in Indonesia that occurred in 1971-1975 as the sample to predict a crisis of Indonesia in 1997 whose results were in general quite. There are several models developed to predict the financial crisis imn Indonesia include: parametric models/econometric models and non parametric, such as signals, the model logitlprobit, hybrid model that includes macro and micro prudential and artificial neural network model. Each using a similar indicator, only the number of indicators (its nodes) are indeed different. From the various models and methods developed, has the performance and results of analysis or early indicators are almost the same/similar and significant at 5% alpha level. It's just an early indicator of the difference and the probability that a crisis will occur when using a forecasting and prediction in outside sample. The performance of each model is adequate in terms of occurancy in predicting the probability of financial crisis in Indonesia. Early indicators are an important contribution in predicting the probability of financial crisis in Indonesia are: real exchange rate, fiscal deficit. the credit ratio to Gross Domestic Product (GOP, world oil prices, growth in export and import, government consumption, domestic Imterest rate spread is the interest rate "The Fed" (The Federal Reserve The Central Bank of The USA).