In recent years, wild boar (Sus scrofa) has been regarded as a typical representative of human-animal conflicts resulted from the population increasing sharply and its consequences of ecological impacts in the mountainous forest ecosystem of southern China. However, few studies can be found on wild boar's habitat selection in southern mountain area so far. In this study, a long-term camera-trapping dataset (from July 2015 to January 2020) obtained from several representative places in Guizhou Plateau were analyzed, and a total of 201 wild boar used plots and 121 random plots were obtained. 1) Vanderploeg and Scavia Selection Index analysis showed that the wild boar preferred the habitat where the slope ≤ 20°, tree canopy coverage >0.8 and herb coverage 0.2-0.4, while avoided the habitat where the slope >25°, herb cover<0.1 and the closest distance from road ≤ 100 m. 2) The Resource Selection Function model of wild boar's habitat selection can be calculated as Logit(P)=3.226-altitude×0.002-slope×0.161 + canopy density of trees×2.078 + shrub average height×0.401 + herb coverage×3.566 + distance from the nearest road×0.001 + distance from the nearest residential area×0.0003, accordingly, the probability of this model was expressed as P=e^logit(p)/(1+e^logit(p)). The prediction accuracy of the model evaluated by the Receiver Operating Characteristic (ROC) can reach to 87.8%, which indicated the wild boar habitat selection behavior can be predicted well. 3) Mann-Whitney U and Chi-Square Test were used to examine the differences in the selection of wild boar for each habitat factor between the used and random plots in different study sites, the factors including topography, vegetation types and human interference also influenced the habitat selection of wild boars. Our findings could furtherly enrich the the ecology of wild boars living in southern China, and would be also a useful reference toward the managements on human-wildlife conflicts and conservation.