The work is devoted to the design of selection method of the type and structure of artificial neural networks (ANN) for restoring the surface distribution of chemical elements (Si, K, Cr, Ti, V, Mn, Fe, Ni, Zr) in soil. A site for the study was chosen in the form of a square with a side of 1 m away from sources of contamination.100 samples of the topsoil (depth 0,05 m) were selected in this area. To restore the surface distribution of chemical elements in the soil using computer modeling, various types and structures of ANN were chosen. For each chemical element, its own ANN was selected and its own estimation of the prediction accuracy was used. Comparison of the concentrations of the surface distribution of chemical elements in the soil, made by different ANNs with known values of concentrations showed that the trained ANNs provide a high prediction accuracy. The proposed approach with cross-validation allows choosing the type and structure of the neural network for an arbitrary site, which is one of the main difficulties in modeling the distribution of chemical elements by the ANN method.