Introduction: Internal dose assessment is usually performed using pre-calculated reference data derived from humanoid anatomical models. Medical Internal Radiation Dose (MIRD) has been the main source of reference data for dose assessment in nuclear medicine for diagnostic and protection objectives. These reference data are calculated assuming uniform distribution of radioisotopes in the organs of some limited number of humanoid models. Practically, however, human anatomy is too much variable that can be represented using a limited number of models. But they are valid in simple conditions they have been derived. GATE is a dedicated code to nuclear medicine and has almost been validated for internal dosimetry application; however, GATE results have never been compared to MIRD reference data. In this study, we constructed the digital form of Snyder phantom and used it with GATE to calculate the doses to the organs from photons of different energies.Methods: The mathematical Snyder phantom sampled digitally and converted to a voxel phantom. Activity was distributed uniformly within the kidneys, liver, lungs, pancreas, spleen and adrenals. The GATE Monte Carlo package was used for calculation of doses in the organs of phantoms. Simulations were performed for gamma photons of 10-1000ke V. Data derived with GATE was then compared to MIRD published data.Results: The results imply a negligible bias between MIRD and GATE data for self-irradiated organs. Bland-Altman analysis shows that the SAF values derived with GATE are on average 0.16% smaller than MIRD values for self-irradiated organs. Data derived for cross-irradiation shows a good linear relationship between SAF values derived with GATE and corresponding MIRD data. The average relative differences between MIRD and GATE data for self-irradiated organs are below 3%. For cross-irradiated organs, the relative difference was 6-10% for photon energy >30 and quite considerable (>25%) for photon energies of 10, 15, 20, and 30keV.Conclusion: Comparison of our data with corresponding MIRD data for cross-irradiated organs showed a high dependency to absolute value of SAF. Correlation between our data and MIRD was high when SAF value was high and decreased as SAF value decreased. As a result the consistency between SAF values derived with GATE and corresponding MIRD data is high when absolute value of MIRD data are greater than 0.01.