Current clinical practice for dose calculations for brachytherapy utilizes the TG-43 formalism where absorbed dose is calculated in a homogeneous water environment. The formalism does not account for the effect of tissue heterogeneities, interseed attenuation, and the finite dimensions of patients causing significant errors in calculated doses for low-energy permanent implant brachytherapy. As an alternative, Monte Carlo (MC) dose calculations model radiation transport and dose deposition in non-water media but have only seen recent application to brachytherapy; issues relating to accurate MC calculations for permanent brachytherapy are unresolved and, in some cases, are completely unexplored. This thesis investigates the accurate application of MC dose calculations for permanent implant breast and lung brachytherapy.MC calculations of permanent implant breast brachytherapy have commonly used a single tissue with a composition averaging that of fibroglandular and adipose tissues. Changes in dose due to segmenting gland and adipose tissues and calcifications are determined and results suggest that averaged tissues produce inaccurate photon energy fluences and thus, in most cases, are unsuitable for calculating doses.The first investigation of patient-specific MC dose calculations for permanent implant I-125 lung brachytherapy is presented where the modelling choices that significantly affect doses are determined and deviations of up to 36% from TG-43 calculated doses are found. CT image artifact reduction and organ-constrained tissue assignment are studied to find improved techniques for modelling patients. MC calculations are used to determine that current lung brachytherapy treatment planning practices are insufficient for producing dose distributions in patients since implant deformation and patient tissues significantly change planning dose distributions. Finally, current and novel radionuclides for permanent implant lung brachytherapy are explored.