[en] Highlights: • Efficiency difference between reference materials and soil samples has been studied using an MCNP model of real HPGe detector. • Elemental compositions of a large number of soil samples and IAEA reference materials were investigated. • Effects of soil elemental composition and gamma energy on sample and reference material efficiency difference were discussed. - Abstract: The relative discrepancies between the gamma spectrometry efficiency of RGU, RGTh, RGK reference materials and some soil samples have been studied using a MCNP model of a real HPGe detector. It has been shown that, in a specified geometry, efficiencies differences depend on the sample elemental composition. The elemental compositions of RGU-1, RGTh-1 reference materials and a soil sample have been evaluated using X-Ray fluorescence (XRF) method and used in the MCNP simulation along with RGK-1 and six other soil samples with different elemental compositions to calculate their efficiencies in different gamma ray energies. To estimate the maximum relative efficiencies differences between soil samples and reference materials, five soil samples with higher attenuation properties were selected from a large data set of soils elemental compositions. The results show that the efficiency differences between soil samples and reference materials are almost ignorable for more than 100 KeV gamma energies. It strongly depends on the sample attenuation factor in the lower energies, so use of a self-attenuation correction is essential for radionuclide counting in low energies gamma rays. Results show about 8 percent discrepancy between RGU and two soil samples efficiencies in 63.2 KeV energy.