The current approach for brachytherapy usually relies on superposition of single sources in a homogeneous liquid water phantom. This technique is fast and practical in the clinical context but not adapted to the present treatment requirements [Beaulieu et al, TG186, Med Phys. 39, 6208 (2012)]. The influence of tissue heterogeneities, inter-seed attenuation and finite patient dimensions, have to be considered. A full Monte-Carlo (MC) code allows computing the Dose Distribution (DD), with a high precision, but is too time consuming to be routinely used in clinical context. This work proposes a new method for accurately calculating the DD within a time compatible with clinical requirements. Our code solves the three dimensional linear Boltzmann transport equation (LBTE). The model is based on a multi-group energy approach combined with a specific angular momentum closure, deduced from the principle of entropy minimization. The code has been recently applied to external radio-therapy and results suggest that the method is as accurate as Monte-Carlo code for the considered test cases [1]. The transport, attenuation and scattering are based on microscopic cross sections in the energy range from a few keV up to Mev’s. All particles, electrons, positrons, photons, primary or secondary are treated simultaneously and computations are performed in full three-dimensional geometry. Our deterministic method differs from existing commercial software [2] by the phase space discretization and the LBTE solving. Our method has been tested by using water phantoms with tissues and density/material inserts. The DD’s computed with our deterministic model have been compared with full three-dimensional converging MC simulations. We demonstrated that the model is capable of reproducing the DD’s with accuracy of a few percent compared to the reference MC calculations. The deterministic model is fast and hence it may be a viable candidate for a future Treatment Planning System. Limitations and possible improvements of the deterministic method are presented. Photon energies are relatively low in brachytherapy and the contribution of scattered photons to the total dose can be of the same order of magnitude as the primary dose, particularly in the presence of inhomogeneities. In addition, for these energies, the photoelectric effect, specifically in the inter seed attenuation, yields deviation from an in-water DD. The proposed method provides a fast, accurate and integrated dose computation without any corrections or adjustments. In addition, the discretization of the space dimension is well suited to problems in radiation therapy where voxel-based geometries of patients are conducted from tomographic images.