Tourism not only brings economic growth to the destination but also has a certain impact on its ecological environment. Existing studies have paid more attention to the pressure of tourism on the ecological environment but paid less attention to the impact of tourism on destination ecosystem service value (ESV). Therefore, based on the theory of tourism society-ecosystem, this study constructs a framework of the impact of tourism on ESV in the destination. Based on the change in land use pattern and the modified ESV equivalent, we calculated the ESV of Shennongjia during 2005—2018 and analyzed its spatio-temporal variation characteristics. Then we identified the influences of tourism factors on ESV, including distance to residential area (DRA), distance to scenic spot (DSS), distance to hotel (DH), and distance to road (DR). Finally, geographically weighted regression(GWR) and boosting regression tree(BRT) models were used to study the influences of tourism factors on ESV and spatial differences. The following conclusions were drawn: (1) the changes in land use pattern in Shennongjia were significant. The areas of cultivated land, forestland, grassland and unused land decreased to 2083.45hm2, 8018.62hm2, 285464.31hm2, 9474.68hm2 and 0hm2, respectively, by 2018. The areas of water and construction land increased. In 2018, it increased to 191.64hm2 and 926.79hm2, respectively. (2) The ESV of Shennongjia showed an increasing trend year by year, increasing from 3.358billion yuan in 2005 to 8.91billion yuan in 2018. From the perspective of ecological service functions, 11 ecosystem service functions accounted for a stable proportion of the ESV. The climate regulation, hydrological regulation, soil conservation and biodiversity accounted for approximately 70% of the value of ecosystem service functions. From the perspective of spatial and temporal evolution, ESV was highly autocorrelated, showing a high distribution pattern in the middle and low boundary. (3) The GWR results showed that DSS and DR had significantly positive influence on ESV, while DH and DRA had significantly negative influence on ESV, and the influence degrees were DSS, DRA, DR and DH, respectively. The BRT results showed that the factors of ESV change in descending order were DRA, DSS, DH and DR. (4) In terms of spatial heterogeneity, GWR could identify the spatial difference and the direction of the effect globally, while BRT could more accurately identify the scope of the main factors affecting the change in ESV.