The sett plays an important role in the badger habitat selection for providing a shelter to resist adverse weather, escape predators, and reproduce. In the context of rapid transformation of agriculture lands to forest plantations and increasing human development activities, understanding the badger sett habitat selection is requisite to conserve badger habitats. Nevertheless, the Asian badger sett habitat selection has rarely been studied in China. From September 2008 to August 2009, we used six randomly selected 8 km line-transects that covers the entire study area to investigate the sett habitat of Asian badger (Meles leucurus) in Xinfeng Forest Farm of Fangzheng Forestry Bureau, Heilongjiang Province, China. We established a plot of 10 × 10 square meters that centers at each of the detected setts and measured 18 variables of the sett habitat (vegetation type, elevation, dominant tree species, tree density, tree height, average diameter at breast height of the nearest tree, distance to the nearest tree, canopy closure, shrub density, shrub height, distance to the nearest shrub, herbage coverage, distance to water resource, distance to cropland, distance to human disturbance, slope, aspect, and slope position). It was found from a total of 55 badger setts that the average size, depth, and obliquity of setts were (27.40±14.30) cm, (84.18±44.07) cm, and (26.36±18.2) degree, respectively. We also used equal number of plots of 10 × 10 square meters (i.e. 55) as controls to collect the same variables of the sett habitat, which were randomly established 500 m away from the detected setts in a random direction. In this paper, the Kolmogorov Smirnov test was used to assess the normality of 14 numerical sett habitat variables. The two-tailed t-test and Mann-Whitey U test were used to compare the sett habitat variables measured from both the sett and control plots. The results reveal that the Asian badger prefers the setts with lower canopy closure and herbage coverage, and higher shrub density as well as with moderate slope, southern aspect, and middle position, closer to shrubs, water, and cropland, far away from trees and human disturbance. Resource selection functions (RSFs) are widely used approach of mathematical modeling for the analysis of presence-absence data to deduce wildlife-habitat relationships. However, the generation of RSFs may be hampered by a wild animal's population size and density. In this study, a log-linear model was fitted to the four sett habitat variables: 246.980-1.059×herbage coverage-0.703×distance to water resource-1.403×slope degree-45.005×aspect. The overall classification accuracy of the model was approximately 93.9% to distinguish the sett habitat from the control habitat. Partly as a result of policies, the Xinfeng Forest Farm is experiencing an increase in agricultural conversion, conversion to tree plantations, and improvement of the road network. These changes may have negative impacts on badgers. Hence, understanding sett habitat selection would enhance conservation of Asian badgers and provide basic data to further guide ecological study.