In the absence of procedures for evaluating the design of brachytherapy (BT) facilities for radiation protection purposes, the methodology used for external beam radiotherapy facilities is often adapted. The purpose of this study is to adapt the NCRP 151 methodology for estimating the air-kerma rate at the door in BT facilities. Such methodology was checked against Monte Carlo (MC) techniques using the code Geant4. Five different facility designs were studied for (192)Ir and (60)Co HDR applications to account for several different bunker layouts.For the estimation of the lead thickness needed at the door, the use of transmission data for the real spectra at the door instead of the ones emitted by (192)Ir and (60)Co will reduce the lead thickness by a factor of five for (192)Ir and ten for (60)Co. This will significantly lighten the door and hence simplify construction and operating requirements for all bunkers.The adaptation proposed in this study to estimate the air-kerma rate at the door depends on the complexity of the maze: it provides good results for bunkers with a maze (i.e. similar to those used for linacs for which the NCRP 151 methodology was developed) but fails for less conventional designs. For those facilities, a specific Monte Carlo study is in order for reasons of safety and cost-effectiveness.