Aim — to study the effect of matrix metalloproteinase-9 on left ventricular remodeling in patients with chronic heart failure after previous myocardial infarction. Material and methods. The study included 118 patients with coronary artery disease and previous myocardial infarction. Blood sampling was performed during coronary angiograph y in 80 patients and prior to artificial circulation onset in 37 cases. Patients were divided into 2 groups: group 1 included 76 patients with end-diastolic volume > 160 ml (EDVb), group 2 — 41 patients with end-diastolic volume 160 ml or less (EDVs). Solid-phase enzyme immunoassay was used to determine matrix metalloproteinase-9 concentration. Multiple logistic regression was applied to identify the relationship between matrix metalloproteinase-9 concentration and echocardiographic parameters. Artificial neural networks (multilayer perceptron) and Poisson’s regression were used to determine the importance of predictors and boundaries of the relationship, respectively. Results. Increase of end-diastolic volume by 1 ml is followed by advanced probability of increased level of matrix metalloproteinase-9 by 8%. Correlation is significant for end-diastolic volume over 193 ml. Reduced ejection fraction by 1% is associated with increased likelihood of augmentation of the level of matrix metalloproteinase-9 by 10%. Association is significant for ejection fraction <53%. Increase of Syntax Score by one point is accompanied by advanced chance of augmentation of the level of matrix metalloproteinase-9 by 4%. Correlation of Syntax Score and concentration of matrix metalloproteinase-9 is permanent. Coronary sinus plasma has the greatest diagnostic significance to determine concentration of matrix metalloproteinase-9.