Probability of Default (PD) is a financial term describing the likelihood of default over a particular time horizon. This concept has attracted a lot of interest ever since the late 1960"s and has been extended to the banking sector to predict probability of failure as well as bank performance ratings. We derive the probability of bankruptcy and bank ratings in a Zimbabwean context based on data between 2009 and 2013, inclusive.We build a model to predict the probability of bank failure twelve months in advance for Zimbabwean banks based on twelve micro factors. Further, we build the corresponding rating model. The empirical analysis revealed that the warning signal so developed produced a robust result with a high prediction accuracy of 92.31% compared to 60% of the Altman"s Z Score model.