Low-carbon behavior of scenic spots has direct influence on coordinated and orderly development of eco-environment and socio-economic system in World Heritage Site. Five level-one indicators and 38 level-two indicators were respectively established to systematically measure the low-carbon behavioral performance (LCBP) of scenic spots in Zhangjiajie, the seat of World Natural Heritage. ANOVA and Bonferroni analysis were applied to compare the LCBP of scenic spots in different groups. The redundancy analysis and Monte Carlo permutation were used to figure out the main driving factors that affect scenic area’s LCBP. Results showed that 32 scenic spots’ LCBP was excellent in general, with a weighted mean value of 3.10. The group mean of perfor-mance in low-carbon design, daily energy conservation, water saving management, waste reduction and low-carbon awareness was 0.49, 0.74, 0.24, 1.51 and 0.11, separately. Huanglong Cave performed the best (4.193) and He Long’s Former Residence performed the worst (2.400) in the 32 scenic spots. The scores of 5A, 4A and 3A or no A scenic spots showed no significant difference in most low-carbon behavior indicators, only 11 indicators reflected significant difference among diffe-rent groups. Pressures from investors, administration committee, tourists and local government were main driving factors for low-carbon behavior of scenic spots.