Because there exist some problems in the use of DOE-2, some cities (such as Shanghai) will adopt DeST-h when they work out local energy efficiency standard. In order to examine the correctness of the software in evaluating the relative energy efficiency of various measures, this paper assumes two model buildings with quite different of shape coefficient, which are built in the form of traditional envelope and improved envelope and are set in twelve Chinese cities with totally different hourly weather data. The cities are Guangzhou, Chongqing, Hangzhou, Chengdu, Nanchang, Nanning, Fuzhou, Guiyang, Shanhhai, Wuhan, Changsha and Beijing. The annual cooling and heating energy consumption of unit building area before and after the improvement of the model building envelope is computed by the DeST-h software. Through comparative research, the author finds out that for the twelve cities, DeST-h can basically forecast the interrelation between the annual energy consumption of cooling and heating and the urban climate. When the DeST-h software is used to evaluate the energy efficiency effect of a particular measure (such as improving envelope, changing shape coefficient and increasing air changes) in one city, it seems that the conclusions are reasonable. However, there exist large differences or even opposite conclusions for some cities when the relative variation rates of annual heating and cooling energy consumption for the improvement of building envelope, the change of shape coefficient and the increase of ventilating rate are compared between different cities, which will make the designers at a loss when they carry out the existing energy efficiency standards. These problems remain to be studied deeply and systematically by software developers, and meanwhile they provide references to people of this occupation.