In order to deal with the accidents in power grid containing a large number of distributed generations, a data mining-based accident response and decision-making method for integrated transmission and distribution grid is proposed. Firstly, massive samples are generated by Monte Carlo simulation and an operation scenario set is established. Secondly, the control strategies with respect to anticipated accidents of different operation scenarios are obtained by optimal service restoration of integrated transmission and distribution grid, and the control strategy knowledge database is established. Finally, the scenario decision-making rule database is established by using the random forest algorithm. When an accident occurs in a real power grid, the control decision corresponded to the practical state of the power grid is obtained by quickly scenario-decision matching. Simulation results on a power grid combining IEEE30 & IEEE33 bus test system by using R software show the effectiveness of the proposed method. This work is supported by National Natural Science Foundation of China (No. 51577049) and Science and Technology Project of State Grid Corporation of China “Research on the key technologies of integrated analysis and fast simulation of main and distribution network with DGs”.