Accuracy forecasting of electric load is important for power system to make plan. A Multiple Factors Addictive(MFA) model is proposed to predict mid-term electric load based on Europe(EUNITE) competition dataset and North American electric dataset. Firstly, MFA considers factors such as temperature, holiday, and week separately to fit functions for electric loads. And then all these fitted functions are added together to a unified function, which is used to make prediction of the electric load. Seven other state-of-art algorithms which are popular in the field are also used to make forecasting. The performances of prediction models are evaluated by using 6 different metrics. Compared with 7 other kinds of different models prediction results, MFA has the advantages of more accurate forecasting performance and faster operational speed, and is simple and easy to understand.