Abstract: In order to use the "big data" of coal mine accidents effectively to analyze the potential laws of the accidents, the causal model of coal mine including factors of environment, equipment and human is proposed in this work. Based on the causal model, the database system of coal mine accident is established. Furthermore, Apriori algorithm is used to mine the causal chain of coal mine accidents in the established database system, and then the maximum frequent item sets are given. Furthermore, the relationship among these causal factors is analyzed and the causal chain of coal mine gas accidents is finally obtained. The results show that the violation of the regulations is the immediate causal factor for coal mine accidents, simultaneously, the inefficient implementation of regulation is the core causal factor for these accidents. Then, the causal chain of coal mine gas accident can be cut down by the way of efficient implementation of regulation and rules. As a result, the occurrence probability of coal mine gas accident can be largely reduced.