In this paper, we identified potential corridors and buffers for connecting all the existing sub-populations of the Yunnan snub-nosed monkey (Rhinopithecus bieti) based on graph theory approach and the principle of the least cost distance and the least home range model for the monkey species. According to the correlation between genetic distance and habitat connectivity of the least cost distance, the threshold for potential dispersal was identified for the sub-populations in each of the five management units. Potential connectivity (PC) index could be then evaluated with the threshold, and the PC index could be used to estimate the impact on the monkey's habitat by agricultural land in terms of landscape connectivity at different scale scenarios. In this study, we found:(1)the threshold for the most optimum least cost distance of dispersal was estimated about 1600 between the monkey sub-populations in each of the management units in a 2.15 km buffer.(2)The improvement of four types the agricultural land situation could promote increasing landscape connectivity.(3)the 15 existing Yunnan snub-nosed monkey sub-populations were divided into 3 different quantities and structural in the patch scale, in scenario C1 and C2,the landscape connectivity was not changed, while in scenario C3 and scenario C4 the landscape connectivity was decreased and adjacent components would merge into one. Thus, the home range of the monkey snub-population would expend, which in turn would be conducive for gene flow. This finding might be significant for conserving the monkey species range-widely.