This paper investigates the use of natural areas, particularly regions in the very early stage of tourism development, specifically to determine how, by taking local residents’ perception of tourism into consideration from the initial stages of development, residents can then be divided into groups with different characteristics, allowing for a better grasp of influencing variables and contributing to the development of different strategies for the different groups. This could be conducive the development of sustainable ecotourism and could aid in avoiding some of the more negative impacts of tourism development. This study takes China's Tiantangzhai scenic area as an example, using the stratified sampling method and multinominal logistic regression model to classify groups of residents based on their perception of tourism, and identifies key factors