The purpose of this research is to assess the accuracy of the Logit model as a tool to predict banlauptcyand to calculate its percentage of accuracy. The sample used in this research is 14 nonfinancial companies listed at Jakarta Stock Exchange over a period of 3 years, that is, from the year 2000 till 2002. The 19 variables used as monetary ratios to predict company bankruptcy derive from Balance Sheet and Statement of Net Income. This research uses Logit Model or binary logistic regression test as used in Ohlson's latest research to discern the most potential variables to predict bankruptcy. The results of this research prove that the Logit Model is sufficiently accurate to predict bankruptcy in companies used in this research. Analysis result accuracy level of 78,57% at the significancy level of 10% with three ratio proved to affect the prediction of bankruptcy, namely total liabilities divided by total assets, current assets divided by total assets, and cash divided by current liabilities.