This article considers the issues of optimizing the system of financial coefficients in early detection of signs of bankruptcy of enterprises. During the study, modern models of bankruptcy prediction, including the discriminant analysis models of Altman, Springate, Taffler and Lis, as well as approaches adapted to the conditions of the national economy, were analyzed. An integrated system of financial stability, liquidity, profitability and business activity coefficients was developed. The forecasting efficiency of financial coefficients was assessed on the example of enterprises of the Republic of Uzbekistan. The results of the study showed that the optimized system of coefficients, taking into account the characteristics of the national economy, allows predicting the risk of bankruptcy with an accuracy of 85-90%. The developed methodology can be used in practice by enterprises, audit organizations and financial analysts.