In nasopharyngeal cancer (NPC) intracavitary brachytherapy, an anatomical dose reference point (in line with that for gynecology work), e.g., at the sphenoid floor, is more precise than the empirical point of 1 cm from the source. However, such increases of the single-source-plan treatment distances may deliver excessive doses inferiorly, to the soft palate. As shielding may help, its efficacy was studied by Monte Carlo simulations in water for 20 and 30 mm diameter spherical NP applicators (representing extremes of sizes for the small NP cavity), with/without lead shielding inferiorly, using a single linear Ir-192, 2 mm steps, equal dwell times for 5 (5DP) and 9 dwell positions (9DP). Dose reductions of the selected points of interest ranged from 1.2% to 40.5% for the 20 mm shielded applicator and a range of 2.9% to 17.9%, for the 30 mm shielded applicator. Dose volume histograms of the "region of interest" (ROI)-a cuboid of 4 x 4 x 0.5 cm3 at the most inferior aspect of the applicator, also differed significantly. The highest doses of the 50% (D50) and 20% (D20) volumes of ROI (for 5DP and 9DP plans) were reduced by 11.9% to 17.9% for the 20 mm applicator and a range of 9.0% to 11.5% for the 30 mm shielded applicator. Doses in unshielded directions were insignificantly changed, for example, with a 20 mm applicator simulated in a 5DP plan, the dose distribution close to the source in the unshielded direction has less than 4% difference at the 50% isodose relative to the dose prescription point. For the 30 mm shielded applicator, despite smaller dose reduction percentages, a more pronounced effective dose reduction was obtained than nominal values when considering radiobiological equivalent doses. Our system was demonstrated to be ready for clinical assessment.