Since the energy consumption of residential buildings is directly affected by the spatial form surrounding the buildings, determining the low-carbon effect of the spatial shape and its optimal scale contributes to realizing the low-carbon transformation of towns and township.Taking Changxing County of Zhejiang Province and Lianjiang County of Fujian Province for example, 49 representative samples of residential buildings are selected in total and included in this research. Based on both GIS analysis and the method of mathematical statistics, with buffer zones built at the interval of 200m within a radius of 1800m around the residential buildings, a continuous-scale research is conducted on the correlation between building energy consumption and the spatial pattern around throughout the year, in the consumption period of the coldest month and in the consumption period of the hottest month.The results showed that:①road density, development density, and floor area ratio were all positively correlated with building energy consumption, but the energy consumption period and scale range corresponding to each index were quite different. ②Water surface ratio and land use mixing degree were common indicators related to building energy consumption, but their positive and negative effects and scale ranges vary in different regions. ③The key form factors that affect the energy consumption of towns in Zhejiang and Fujian during the monthly energy consumption period with the highest temperature were 1000 m- road density, 800 m- water surface ratio, and 1600 m- open space rate; The key form factor that affects the energy consumption during the energy consumption period of the lowest temperature month in Lianjiang is 800 m-water surface ratio. ④In the selection of carbon control unit, it is necessary for Changxing County to focus on the radius range of 1000 -1200 m, and Lianjiang County to focus on the radius range of 800 -1000 m and 1400 m.⑤ Characteristic indexes of towns in counties in Zhejiang and Fujian were the rate of open space, which was negatively correlated to building energy consumption in the consumption period of the hottest month; characteristic indexes of Changxing County are floor area ratio and waterfront density, which were positively and negatively correlated to building energy consumption throughout the year, in the consumption period of the coldest month and in the consumption period of the hottest month respectively. On such a basis, 3 proposals concerning low carbon urban planning are made herein, which may serve as a theoretical basis for the low carbon transition of county towns and energy saving and emission reduction in residential buildings in Zhejiang Province and Fujian Province.