The MCNPX code was used to calculate the TG-43U1 recommended parameters in water and prostate tissue in order to quantify the dosimetric impact in 30 patients treated with I-125 prostate implants when replacing the TG-43U1 formalism parameters calculated in water by a prostate-like medium in the planning system (PS) and to evaluate the uncertainties associated with Monte Carlo (MC) calculations. The prostate density was obtained from the CT of 100 patients with prostate cancer. The deviations between our results for water and the TG-43U1 consensus dataset values were -2.6\% for prostate V-100, -13.0\% for V-150, and -5.8\% for D-90: -2.0\% for rectum V-100, and -5.1\% for D-0.1; -5.0\% for urethra D-10, and -5.1\% for D-30. The same differences between our water and prostate results were all under 0.3\%. Uncertainties estimations were up to 2.9\% for the g(L)(r) function, 13.4\% for the F(r,theta) function and 7.0\% for Lambda, mainly due to seed geometry uncertainties. Uncertainties in extracting the TG-43U1 parameters in the MC simulations as well as in the literature comparison are of the same order of magnitude as the differences between dose distributions computed for water and prostate-like medium. The selection of the parameters for the PS should be done carefully, as it may considerably affect the dose distributions. The seeds internal geometry uncertainties are a major limiting factor in the MC parameters deduction. (C) 2014 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica.