The efficiency of airlines carbon emission reduction is one of the core elements for us to response to the EU ETS and international climate change negotiations. In terms of the fact that carbon emission reduction in air transport is a multi-input and multi-output problem, a method of data envelopment analysis (DEA), which can scientifically evaluate the results of airlines carbon emissions in our country, is put forward. Evaluation indexes system is selected by analyzing the factors of evaluation by building model of airlines carbon emissions by DEA, the linear programming model can be solved and the result of the relative ordering will be got. Finally, it is analyzed the numerical example of relative effectiveness, and the relative efficiency as well as the input redundancy and output deficiency of these airlines are obtained. The result shows that the method is an effective algorithm, which can be well applied to the study of evaluation on airlines carbon emissions and provide decision support to energy conservation and emissions reduction for airlines.