The aim of the research is to measure the role of financial analysis in predicting bank failures, with the aim of distinguishing between troubled and non-performing banks. Finance that can be used to build a model to predict financial failure.The study used a statistical method that relied on Multiple Linear Discriminant Analysis using the Stepwise method, which works to delete the financial ratios with low (non-significant) impact.A set of financial ratios for the study sample banks, which consisted of six banks, three of which are non-performing and the other non-performing, were calculated for the period (from 2010 to 2017), where the following model was reached: ???? = −0.001(????4) + 0.038(????12) + 0.002(????29) − 1.074 The model was applied to the study sample banks, and in light of the results, the banks were reclassified into two groups, troubled and non-performing, with a matching rate of 97.9%.Based on the foregoing, the study recommended the need to use the model and apply it in banks in order to predict financial failure before it occurs so that the banking administration can take measures and procedures that would address the imbalances before they escalate.