Self-absorption effect has a significant impact on gamma spectrometry measurement for the highly precise determination of the radionuclides activity. Among the main factors affecting the self-absorption effect especially at low gamma energies is the matrix composition of the sample. For this purpose, we studied the matrix composition effect on the self-absorption correction for plant and soil matrices using a Monte Carlo simulation model, composed of High Purity Germanium (HPGe) detector and a cylindrical geometry source. The simulated configuration model was validated by comparison of simulated full energy peak efficiency (FEPE) values with the experimental ones using a standard radioactive solution, covering large interval gamma energies in the range of 59.54–1836.12 keV. The results show that the FEPE values of low gamma energies are affected by the matrix composition, where in the soil matrix the FEPEs decreased by 36% and only by 1% in the plant matrix. The Monte Carlo results were applied in order to accurately calculate the transfer factor of the environmental radionuclides 226Ra, 232Th and 40K from soils to several crops, mustard, artichoke and fennel through their specific activities.