The aim of this work was to investigate the applicability of MR-based polymer gel dosimetry to measure the absorbed dose distribution at short distance from an iridium-192 brachytherapy point source. In this paper, different methodological problems that may result in significant errors in the measured dose distribution are discussed. First of all the extent to which physicochemical mechanisms alter the dose response is discussed. The most important among these are the oxygen permeability of the catheter material and monomer-diffusion-related effects during irradiation. The effect of oxygen on the dose-R2 curve has been determined quantitatively and an oxygen map is performed using a well-defined external irradiation beam. The effect of diffusion of monomer during irradiation has been simulated. Another contribution of errors is related to magnetic susceptibility differences between the catheter and the gel during scanning the irradiated gel. The magnetic field distortion has been mapped by using both an experimental MRI technique and by simulation. Moreover, in constructing a dose-versus-distance curve by polar averaging, the sensitivity to the exact distance between source and point of measurement and to partial volume effects is illustrated. An optimization program is proposed to determine the location of the source on a sub-pixel scale.