The problem of the gas emission of the working face was characterized by small samples,nonlinear,and it was affected by complex factors. Using the genetic-least squares support vector regression algorithm to predict the gas emission could avoid the qualitative analysis limitations and effectively improve the accuracy of the forecast,because it was a quantitative method for analysis. First,the model used the genetic algorithms to train and optimize the least squares support vector regression parameters,and then used the genetic-least squares support vector regression model to predict the amount of gas emission of test samples. The test results show that: the genetic algorithm-least squares support vector regression model has a higher reliability and accuracy,compared to the predicted values of the support vector regression and the least squares support vector regression.