A lot of factors somehow bring disequilibrium to the development of city areas， such as area size， service year， area function and location， etc. But few of these factors have been included in the classification and subarea methods. Based on the variable data in practice， city areas may be classified into five kinds: old areas with known historical loads; old areas with only service years being known; old areas of no data; new areas with known service years and new areas of unknown service years， which are respectively forecasted using different approaches. Of all the methods， the template method and merit indexes and weight coefficients method are employed to handle old areas with only service years known and new areas of unknown service years， respectively. The applicability and feasibility of the method are demonstrated with a case study.