Abstract: In order to effectively assess the risk of gas explosion in coal mines, a gas explosion risk assessment method based on ISM-BN is proposed. Firstly, 65 typical gas explosion accident cases were coded at level 3 using grounded theory, and 13 risk factors of gas explosion were extracted; secondly, the Bayesian Network structure is established by interpreting the structure model, and the cloud model is introduced to calculate the prior probability and conditional probability of Bayesian Network nodes, and the probability value is imported into GeNie software for Bayesian Network parameter learning to evaluate the possibility of coal mine gas explosion; finally, the sensitivity analysis and diagnostic reasoning of Bayesian Network model are carried out to analyze the key causative factors of coal mine gas explosion, and put forward corresponding countermeasures. Taking a mine in Shanxi Province as an example to apply the evaluation method, the results show that, in addition to the two necessary conditions of gas explosion, gas overrun and ignition source, coal dust explosion, ventilation system defects, gas emission, electromechanical equipment sparks, coal spontaneous combustion, equipment explosion loss, illegal blasting and other factors are the key factors causing coal mine gas explosion. The underlying reasons for these key causative factors are management factors, and the evaluation results are consistent with the actual situation.