As the depth of underground resource extraction continues to extend, the operating conditions in deep mines are becoming increasingly harsh, and high temperatures and humidity are now posing a huge threat to the health of workers and mine safety production. In an attempt to assist mine management in assessing the riskiness of the extreme hot and humid environment of underground mining, we proposed a multi-indicator grading model. The model was developed around eight factors that affect high heat and humidity environments, using the uncertainty measurement theory. In the evaluation process, the indicator weights were determined by the entropy weighting method, and then the multi-indicator measure evaluation matrix was derived. The confidence recognition criterion was used to determine the grade, so as to obtain the evaluation results of thermal humidity risk. A ranking of the riskiness importance of the evaluation objects was also achieved. Finally, the model was used to analyse and assess the risk rating of thermal humidity at 11 working faces in a gold mine. The results show that the evaluation results are consistent with the real results and this approach can be applied in practical engineering.