The biodiversity conservation priority areas are the important areas designated by China to strengthen biodiversity protection and supervision. At present, some priority areas have successively carried out investigation and evaluation of biodiversity background data. However, the full coverage of all counties or grids in the survey area is neither realistic nor necessary due to the limited funding, time and other conditions. Given that, the sampling design at the regional scale becomes a key issue that needs to be resolved. Taking the biodiversity conservation priority area in the Wuling Mountains as an example, we proposed a scientific and feasible sampling plan by combining hierarchical clustering and systematic sampling. Simultaneously, the systematic integrity of historical survey data and spatial attribute for conservation was also considered. We firstly grouped all counties in the study area into five categories, then selected 2 key counties for each category. According to the above principles, we obtained 6 counties that once investigated scientifically and 4 counties with a relative lack of survey data. Finally, 76 key survey grids (10 km×10 km) were selected, including 36 biodiversity enrichment grids and 40 human disturbance grids outside the protected areas, with a sampling ratio of 11.09%. The average survey funding for each grid was 25,200 RMB. By the stratified sampling and emphasizing on key areas, the surveys and assessments become more target-oriented. The sampling strategy improved the basic survey system of conservation priority areas, strengthened the capacity of project management, and provided a paradigm for other conservation priority areas.