The purpose of this study was to determine differences in predictions of bankruptcy analysis for food and beverage companies in the 2016 - 2018 period used the Altman Z-Score, Zmijewski, Grover, and Springate models and find out which of the most accurate methods of predicting the potential bankruptcy of food and beverage companies. The study population is the food and beverage sub-sector companies in 2016 - 2018. The sampling technique used purposive sampling. Data analysis techniques used the calculation of financial ratios to data used financial ratios, descriptive analysis, the calculation of each model, and calculate the level of accuracy with the help of Microsoft Excel. The analysis shows that there are differences in bankruptcy predictions between the Altman Z-Score, Zmijewski, Grover and Springate models. The analysis model with the highest level of accuracy achieved by the Zmijewski model, then followed by the Altman Z-Score Grover model and finally the Springate model.