The International Commission on Radiation Units and Measurements (ICRU) recommends the use of personal dose equivalent, H_p(d) as operational quantity for personal monitoring (ICRU, 1992; ICRU,1993). Depths of 10 mm and 0.07 mm are recommended for strongly and weakly penetrating radiation, respectively. Since the whole-body personal dosimeter is worn on the trunk, the backscattering contribution from the body must be considered. The backscatter factor (BSF) is individual for each dosimeter and it is influenced by the photon energy, the detector material, the irradiation geometry, and the phantom properties (dimensions and material) (Singh et al, 2013). In practice, a phantom is recommended for the calibration of dosimeters used for the determination of H_p(d) at the defined depth in the human body. The calibration phantom allows to reasonably reproduce the backscattering conditions of the part of the body where it is used. The ICRU (1992) recommends the use of a 30 x 30 x 15 cm³ PMMA slab phantom, while the International Standardization Organization (ISO, 1999) recommends a water-filled phantom, with the dosimeter placed on the front face of the phantom. This work aims to study the impact of using different phantoms on the backscattering and on the response of a prototype of a new personal dosimeter. Monte Carlo simulation with the PENELOPE package was used to calculate the BSF for the ICRU and for the ISO phantom. Determination of the BSF over the photon energy range of 24 keV to 662 keV was performed. The system personal dosimeter is designed as three regions based on a semi-cylindrical 3D-printed badge accommodating three thermoluminescent (TL) pellets of CaSO₄:Tm plus polytetrafluoroethylene (PTFE). The pellets are semi-spherical shaped (5 mm-diameter and 0.9mm-maximum height) with a flat surface to be used with the usual flat planchet of a TL reader. One pellet is shielded/filtered by the badge material (polylactic acid, PLA); one pellet is plus polytetrafluoroethylene (PTFE) shielded by the badge material containingand by a copper plus lead filter; and one pellet is positioned in an “open window”, being filtered only by a mylar sheet. The simulated BSFs show differences of less than 7% for both phantoms when photon beams are considered. Even though differences were found for the use of different phantom, those are within the accepted uncertainty in dose assessment for radiation protection measurements.