Abstract: Metro stations are important lifeline projects in cities, but the number of disaster emergencies in its operation phase is increasing, and the complexity and difficulty of emergency decision-making is also gradually increasing.To improve the scientific and targeted emergency decision-making of subway stations when facing sudden fires and other prominent issues, this paper proposes a subway station fire emergency decision-making method based on ontology theory and knowledge reasoning.The ontology is used to analyze and construct the ontology model of emergency casesand the weight of subway fire feature attributes is calculated based on random forest.A digital emergency plan rule base is established based on SWRL rules, and the intelligent generation of emergency decision-making scheme is realized through case-based reasoning, rules, and ontology-based reasoning machine methods.Finally, a metro station is taken as an example to verify the feasibility of the method.The results show that Case #5 has the greatest similarity, with a similarity value of 0.85,and the reasoning of emergency measures for metro station personnel is realized by using SWRL rules.The method based on ontology and knowledge reasoning can generate the fire emergency response decision-making method of metro stations scientifically, quickly and effectively, which enhances the reuse of knowledge in the field of metro operation, and provides technical support for scientific emergency response and intelligent emergency response to emergencies.