Steel industry is one of the pivotal industries supporting the development of the economy, but it is also characterized by high energy consumption and high emissions, making it a source of energy consumption and carbon emissions. There are currently few research on the forecast of CO 2 emissions of industrial firms, particularly the steel industry, because of the absence of monitoring data. For this reason, this study suggests a technique for predicting carbon emissions from the steel sector based on the electric-carbon correlation model. First, the calculation method of CO 2 emissions in the steel industry is proposed based on the production process of steel. Second, the basic form of the electric-carbon correlation model of the iron and steel sector is generated by choosing electricity consumption as the influencing factor of CO 2 emission. Finally, relevant variables are selected to fit the support vector regression (SVR) to the data of the steel industry to construct a prediction model of CO 2 emissions in the steel industry. The accuracy of the electric-carbon correlation model and the CO 2 emission prediction method in the steel industry are confirmed by the example analysis.