European noise directives advise to apply corrections when measuring and modelling noise levels close to a building in the aim of excluding the contribution of noise reflection against the façade. The advised -3 dB correction is still subject to discussion. In order to investigate the needed correction for a household exposure studies, a high definition noise model was used to estimate noise levels at 10,394 inhabitable buildings. Three buffers were used to sample area surrounding façades of buildings. The surfaces were defined between the following distances: i) 0 and 2m, ii) 0 and 6m, iii) 2 and 6m. No differences between the distribution structures were observed. Mean noise levels do not differ significantly between the buffers methods (respectively i) 49.6±6.7 dB, ii) 49.7±6.7 dB, iii) 49.8±6.7 dB ; p<0.01). Same observation can be made for maximum noise levels (respectively i) 52.0±7.2 dB, ii) 52.5±7.2 dB, iii) 52.4±7.2 dB ; p<0.01). These results show no or light differences between indices computed by the three sampling methods. They are in favour of no or at least a low correction value to deal with the contribution of noise reflection against the façade of a building.