Abstract: Based on the natural ecological factors, such as climate, soil and topography, combined with Oyama sinensis distribution sample data, the Maximum Entropy Model(MaxEnt) optimized in R language Kuenm package was used to predict the potentially suitable distribution areas of Oyama sinensis. On this basis, according to the human disturbance factor calculated by the entropy assignment method, combined with the Marxan model, we identified the priority conservation areas for Oyama sinensis and explored the conservation vacancy areas for Oyama sinensis. The results showed that the AUC value of the optimized MaxEnt model was 0.969, indicating that the optimized MaxEnt model predicted extremely well. The simulation results of the optimized MaxEnt model indicated that, the potentially suitable areas for Oyama sinensis were roughly banded from Jinyang County, Liangshan Yi Autonomous Prefecture, to Nanjiang County, Bazhong City. The main natural ecological factors affecting it distribution were the annual difference in temperature, the precipitation in the driest month, the geological hazards, the altitude, the seasonal changes in temperature and the total exchangeable bases in the topsoil. Human disturbance factors were taken as the cost of protection, and the Marxan model was applied to identify the conservation vacancies of Oyama sinensis mainly located in the Tianquan River Nature Reserve, Gongga Mountain Nature Reserve, and the areas near the Giant Panda National Parks, Erlang Mountain National Forest Parks, and Hailuogou National Forest Park.