Introduction. The study of the influence of chemical composition of materials on their mechanical properties is most often realized through modeling. This approach makes it possible to establish a one-to-one correspondence between composition and properties in the form of the obtained regression equations. The estimation of the magnitude of the coefficients of the equations allows us to determine the degree of influence of chemical composition elements of the materials on mechanical properties. It is therefore proposed to use this approach to evaluate the effect of chemical composition of S420M on its strength. Materials and methodology. The chemical composition and the tensile strength в of the S420M structural steel were varied within the limits of the ДСТУ EN 10025-4 standard for metal products. Using a technique of experiment planning, a fractional replica of the 2 4 experiment matrix was implemented. The results of the experiment. Multiparametric regression equations for the estimation of the tensile strength depending on the percentage of carbon, silicon, manganese, phosphorus, sulfur, chromium, nickel, molybdenum, niobium and vanadium were obtained. The greatest impact on the tensile strength, depending on the percentage, is 125,25 carbon, 29,50  manganese, 29,50  chromium, 24,50  vanadium and 16,50  molybdenum. The calculation of the effect of the elements of the chemical composition on the strength was carried out on the basis of the analysis of the coefficients of the obtained mathematical model. Thus, the approach to the prediction of the strength limit of S420M steel is implemented, which allows to control its performance during production by changing the chemical composition. The performance of the model is confirmed by the statistics of Fisher and Kochren. Conclusions. On the basis of the analysis of the influence of the chemical composition of S420M steel on its ultimate