Abstract: In order to improve the accuracy of risk prediction for coal mine gas explosion, proposes a risk prediction method for gas explosion that combines Fault Tree Analysis (FTA) and Analytic Hierarchy Process (AHP) - Case-Based Reasoning (CBR). The core innovation of this method lies in the construction of "coupling mechanism between fault tree-FTA and AHP-CBR" and "two-level index retrieval" model. 12 items of key disaster-causing factors are identified by using FTA, combined with AHP dynamic weight determination and CBR two-level index retrieval mechanism, accurate prediction of risk occurrence and hidden danger and influence scope is achieved. In instance verification, the accuracy rates reach 100%, 95%, and 100% respectively, which provides an effective prediction tool for gas explosion risk prevention and control in coal mines.