In order to establish a rapid prediction method for the intensity in the meizoseismal area after the earthquake,the paper collects many cases of earthquakes with M ith before 2013. Taking the magnitude and focal depth as input parameters, and the seismic intensity in meizoseismal area as the output parameters, the paper establishes an ELM prediction model and analyzes the information entropy and weights of the magnitude and focal depth. Compared with the traditional model,the prediction accuracy of this model is improved by about 20 %.The main conclusions are as follows:there is a positive correlation between magnitude and focal depth,and a negative correlation between magnitude and seismic intensity. The information entropy of magnitude is higher than that of focal depth, which suggests that its variation degree and information content are less, and influence on seismic intensity is smaller. The effort to be done is to seek prediction means which is suitable for local area and based on multiple parameters.