Background. Decompensation of chronic heart failure (CHF) is one of the most common reasons for hospitalization of older patients. Often this diagnosis at the pre-hospital stage is established unreasonably and subsequently rejected at different stages of patient hospitalization. For primary diagnosis of decompensation of chronic heart failure at the pre-hospital stage or in the emergency room it is necessary to have a screening test that allows to quickly determine the presence of decompensation of the disease. General clinical blood analysis, is one of the most accessible methods of laboratory diagnostics in real clinical practice. Taking into account the growing amount of data received by doctors during patient examination, which leads to an increased workload on medical workers at all stages of care, there is a need to use various machine learning methods to increase the efficiency of processing and interpretation of the provided diagnostic information. Purpose of the study. The aim of the study is to improve the quality of diagnostics of chronic heart failure decompensation during primary examination in elderly patients on the basis of hematologic indices using machine learning algorithms. Method. A prospective pilot study was conducted. A total of 101 patients (34 men, 67 women) were examined. The mean age of the patients was 74 (72;76) years. Inclusion criteria were age over 18 years, preliminary diagnosis "I50.0 - Congestive Heart Failure". All patients underwent a routine examination during hospitalization for a general clinical blood test on an automatic hematological 5-diff analyzer. Results. As a result of application of deep learning method it was possible to reach the area under ROC-curve=0,8077 on the test sample when assessing the quality of diagnosis of decompensation of CHF in patients of different age groups. The achieved quality of primary diagnosis can be considered acceptable and promising for further accumulation of the database with the purpose of additional training of the developed algorithm and increasing the accuracy characteristics of diagnosis of chronic heart failure decompensation. Conclusion. The achieved quality of primary diagnosis can be considered acceptable and promising for further accumulation of the database with the purpose of additional training of the developed algorithm and increasing the accuracy characteristics of diagnosis of chronic heart failure decompensation.