In this paper, by the application of model for risk assessment of business enterprises the results which evaluated the financial stability and the likelihood of bankruptcy were obtained. For the calculation and analysis of models, financial indicators and vertical and horizontal analyzes, data for the period 2014-2015 were used, which were published on the pages of the Financial Agency of the Republic of Croatia (FINA). The analysis was carried out on the data of Agrokor, a company which one year before the collapse was the largest Croatian private company and one of the strongest companies in Southeast Europe with a consolidated total revenue of 49 billion HRK and nearly 60,000 employees. The aim of this paper is to determine whether the risk assessment models can predict the company's financial failure, two years before it occurs, on the basis of publicly disclosed financial statements. The analysis was carried out using models developed by Beaver, Zmjiewski, Springate, Fulmer, Kralicek, Chesser, Ohlson and Altman. A total of 13 models were analyzed, and all of the models and financial analysis pointed to the same answer. Two years before the financial breakdown Agrokor Ltd. was already a heavily indebted company which had problems with liquidity and inefficient use of property, which is why all of the models predicted the inability to repay loans, financial problems and bankruptcy. The risk of bankruptcy could have been reduced by acting on the items that in the models affect the risk increase. One way would be to dis-invest and pay back the debts, but the Agrokor administration made the opposite decision and carried out the acquisition of Mercator, which resulted in a financial breakdown of the company. The research results confirm the thesis that risk assessment models successfully predict the financial difficulties in which Agrokor found himself, two years before they were came true, on the basis of publicly disclosed financial statements.