Traditional analytical methods for calculating and designing of grounding devices (GDs) in heterogeneous soils can lead to results that do not correspond to those obtained in practice. The empirical coefficients used in calculations of grounding systems given in various literature sources do not always give an explicable and significant discrepancy with the same initial data. In this paper, some well-known methods for calculating the spreading resistance of GDs are considered and the obtained results are compared with experimental data. It is shown that the recommendations and algorithms for calculating the resistance of GDs, presented in the well-known reference literature and regulatory documentation, do not give full and correct description for grounding devices installed in heterogeneous soil. In particular, it is shown that such a factor as the proportional ratio of soils with different resistivity practically does not affect the final result. This fact may mislead specialists, since the results obtained may differ significantly from what is observed in practice after the installation of GDs. The study proposes a calculation method and shows a fairly good convergence of the results with experimental data, and defines a further direction in optimizing calculation methods GDs.