Abstract: Abstract：Scenic accessibility is an important basis for the development of tourism. On the basis of systematic analysis of the scenic spot accessibility’s subject and object connotation, and considering the subjective priority of tourists to scenic spot selection caused by the difference of tourism traffic service level, a scenic spot accessibility measurement model is constructed from the perspective of scenic spots’ accessibility of external traffic and urban traffic transfer. Take 76 scenic spots above 4A level in Yunnan Province as an example, and put forward some optimized strategies of scenic accessibility. The results indicate that: (1) Spatial distribution characteristics of external traffic accessibility is greatly different, about 60% scenic spots’ external traffic system development lags behind, and they distribute in the county discretely; intra-urban traffic transfer efficiency of scenic spots is great, the average transfer time is 0.472 h, some scenic spots’ level don’t match their accessibility level. (2) The accessibility level of scenic spots is overall poor-about 33% scenic spots’ accessibility level is average or worse, and the spatial distribution of accessibility has a high or low value concentration phenomenon. The hot spots of accessibility have obvious decreasing laws from east to west, and external traffic accessibility of scenic spot is the main reason for that. (3) We should take the "tourism point axis structure" as the development concept, give full play to the core node function, and improve the construction of tourist transport corridors by connecting important dots into lines to increase the external accessibility of the scenic spot. Enhance the connection between scenic spots and major transport hub, and achieve the coordination of external traffic accessibility and intra-urban traffic transfer efficiency. The construction of recreational roads according to local conditions can ease the intra-urban transfer disadvantages of natural scenery scenic spots.