Abstract: A pre-warning method of gas concentration based on correlation analysis of monitoring data was proposed. Firstly, statistical characteristic of mine gas monitoring data is analyzed, intrinsic correlation characteristic of time series consisted by the gas monitoring data is researched by use of grey correlation analysis method. Further, pre-warning index and pre-warning threshold are determined based on real-time monitoring data. Then abnormal situation of gas concentration is analyzed and pre-warning level is divided, so as to realize dynamic and quantitative pre-warning. The case analysis shows that the method is applicability for gas concentration pre-warning of single monitoring site, which can provide effective decision basis for daily safety management of coal mine.