The low energy model of the GEANT4 Monte Carlo toolkit was used to simulate the energy response of a T.P.A. Mk-II ionization chamber under a variety of different conditions. The sample position resulting in the maximum response along the axial direction of the chamber was obtained. The parameters of the simulation were chosen to account for the maximum effect of the particle backscattering and the setting of most suitable values for the production thresholds and the energy cuts in the GEANT4 Monte Carlo code. The chamber response for different compositions of detector elements was also studied. The simulated radioactive source was a glass ampoule containing 3.6 ml of the radionuclide in an aqueous solution. The energy response of the chamber at the maximum response was obtained for simulations for 60Co, 22Na and 59Fe nuclides. Verification of the simulated response was obtained using experimental measurements with radioactive sources. The simulated results were in good agreement with the experimentally measured data to within 0.04-2.0%. In the energy range below 200 keV the response curve was complex due to the increase of photoelectric cross sections of the chamber materials. Effects due to backscattering occur at boundaries between chamber elements and also become significant at sites of lead shielding at photon energies above 700 keV. The chamber response for different compositions of detector elements was also studied. The response of the chamber depended highly on the energies of emitting particles, source position and materials used in electrodes and thimble wall.