Background. According to the literature data, acute coronary syndrome (ACS) in 2-20 % of cases is combined with atrial fibrillation (AF). According to the current guidelines of the European Society of Cardiology (ESC), patients with coexisting AF and ACS should receive dual antiplatelet therapy for the prevention of recurrent cardiovascular events and anticoagulant therapy for the prevention of thromboembolic complications. However, this combination is fraught with the development of hemorrhagic syndrome.Aim. To develop a model and software module for predicting possible bleeding in patients with ACS combined with AF taking three-component antithrombotic therapy.Materials and Methods. To build prognostic models for the development of hemorrhagic syndrome, a statistical method was used for classification trees and the neural network procedure implemented in the STATISTICA package. To build prognostic models, a sample was used consisting of 201 patients with a combination of ACS and AF with and without fatal outcome, the state of which was described by 42 quantitative and qualitative clinical indicators. The control group included 205 patients with ACS and intact sinus rhythm.Results. To identify predictors of predictive models of the possible development of hemorrhagic syndrome in patients with triple antithrombotic therapy, the Spearman correlation coefficient was used. The study of correlations allowed to reveal clinical indicators – predictors of prognostic models. After analyzing the predictive ability of the developed models, a software module was created in the Microsoft Visual C # 2015 programming environment that allows determining the possibility of hemorrhagic syndrome in patients with a combination of ACS and AF using classification trees and neural networks.Сonclusion. A classification model and a software module were developed to predict possible bleeding in patients taking three-component antithrombotic therapy. Models contain both quantitative and qualitative (categorical) clinical indicators. The current level of development of data analysis technologies opens up broad horizons for medicine in solving problems on real medical data, without translating them into scoring risk scales, including prediction of the diagnosis of the disease, stage of the disease, treatment outcome, possible complications, etc. High reliability of such systems can be provided by large volumes of medical data accumulated on servers.