This work present to the mathematical model in the form of ANN, intended for projecting concrete compressive strength. Input data was classified according to the type of component material and its content in concrete mix (cement contents, coarse aggregate, fine aggregate, water and admixtures). In order to determine mathematical model, a multilayer, one-way perceptron network was used, recursion network with sigmoidal neurons. The model assumes that neurons are gathered in some layers (one input layer, hidden layers and one output layer). The conducted cross-section of the influence of variables parameters values (learning constant-α and momentum valuesη) on the accuracy of representation of compressive strength was analyses. Assessment criterion was assumed taking into consideration the lowest mistake level and 100% compliance. According to the obtained analysis results ANN was assumed the best representing network for constant value of momentum 0,3, learning constant of 0,05 and 6 neurons in a hidden layer. Very good coincidence of component models with experiment results was achieved. At testing stage, the coincidence was achieved at the level of 99.74%, in case of the assumed network structure. During model verification by means of experimental results, the average coincidence was 99.83%.