Humidity correction on atmospheric extinction coefficient is an important component of aerosol hygroscopicity research, which is also a key technical link for determining the mass concentration of near-ground particulate matter retrieved by satellite aerosol optical depth(AOD). Based on hourly PM_2.5 mass concentration, atmospheric visibility, and relative humidity (RH) data collected in Chengdu from January to December 2016, the statistical relationship between the atmospheric extinction coefficient and particulate matter mass concentration, as well as its response characteristics with RH variation, are discussed in detail. The atmospheric extinction coefficient per unit mass follows a log-normal distribution when RH is less than 90%, and the shape and scale parameters of the distribution function exhibit a fluctuating growth as RH increases. Second, based on the log-normal distribution parameter of atmospheric extinction coefficient per unit mass under dry environmental conditions (RH ≤ 40%), the effect of humidity change on the probability distribution parameter of atmospheric extinction coefficient per unit mass is eliminated via mathematical transformation. Consequently, the principle and flow chart of humidity correction are proposed for the atmospheric extinction coefficient. Finally, the applicability based on the principle demonstrates that the correlation coefficient of PM_2.5 mass concentration calculated by corrected atmospheric extinction coefficient and actual PM_2.5 mass concentration is up to 0.90, significantly improving the corresponding humidity correction result using aerosol scattering hygroscopic growth factor method.