As the hot and humid environments in deep underground spaces deteriorate with increasing mining depth, there is an increased demand to accurately assess worker heat stress levels in underground environments characterized by high relative humidity and air velocity. A modified predicted heat strain (mPHS) model was proposed through the correction factors of air velocity and relative humidity for the clothing thermal insulation and vapor resistance. The predictive performances of the mPHS model for mean skin temperature, sweat loss, and core temperature were validated in low, moderate, and high metabolic rates. The model's guidances for deep underground environments were discussed, and the maximum allowable exposure times (MAET) for 168 conditions were provided. The results demonstrated that the mPHS model exhibited superior predictive performance within 60-320 W/m