Abstract: Taking mine gas emission prediction as the main purpose, in view of the predict difficulties of large amount of influence factors, complex coupling mechanism and variation laws, we present an integrated analysis based on empirical mode comprehensive analysis of decomposition (EMD) and auto-regressive and moving average model (ARMA) time series, and use disposed time-sequence processing of mine gas emission to achieve specific forecast. According to the different characteristics of each sample, we select suitable model to establish EMD-ARMA forecasting model for the forecast of mine gas emission and trend.