Representative Sb tailings from the northwest and southwest regions were selected to investigate the migration characteristics of Sb and arsenic (As), as well as changes in microbial community under both covered and exposed conditions. The results showed that the total concentrations of Sb-As in the Qinglong tailings were significantly higher than those in the Xihe soil-covered tailings, with Sb levels being 21.90~29.35 times higher and As levels 20.23~39.26 times higher. The index of Geo-accumulation (I_geo) and the Ratio of secondary phase to primary phase (RSP) indicated that the Qinglong tailings Sb-As contamination index was greater than 3, which belongs to severe contamination, the exposure of Sb-As thus poses a considerable threat to the soil environment. In contrast, the I_geo index for Sb-As in the Xihe soil-covered tailings were both <3, and RSP values was <2, indicating a relatively low level of contamination. Using the Wenzel sequential extraction method, the distribution of Sb-As fractions was found to follow the order: amorphous iron-aluminum oxide-bound fraction (F3)>residual fraction (F5)>crystalline iron-aluminum oxide-bound fraction (F4)>specifically adsorbed fraction (F2)>non-specifically adsorbed fraction (F1). F3 accounted for 36.91% and 37.00% of the total Sb-As contents, respectively. The bioavailable forms of Sb [Sb(Bio)] and As [As (Bio)] were significantly higher in the Qinglong tailings than in the Xihe soil-covered tailings. In Qinglong, Sb-As primarily occurred in the F3fraction, whereas in Xihe, they were mainly associated with the F5fraction. Principal coordinates analysis(PCoA) revealed that bacterial community structure varied across different environmental media. The dominant bacterial phyla in the Qinglong tailings were Chloroflexi and Firmicutes, with Kaistobacter, Pseudomonas, and Rhodoplanes as the predominant genera. In the Xihe soil-covered tailings, Gemmatimonadetes and Bacteroidetes were significant components, and Thiobacillus, Kaistobacter, and Sphingomonas were the dominant genera. Thiobacillus, Pseudomonas, Bacillus, Geobacter, and Sphingomonas were identified as key bacterial taxa enriched in long-term Sb-As contaminated environments, indicating their potential for application in in-situ ecological restoration of Sb tailings. The functional prediction using PICRUSt2 and Partial least squares discriminant analysis(PLS-DA)suggested that bacterial functional changes were primarily characterized by the enrichment of specific metabolic functions, accounting for approximately 40.66% of the total predicted functions. Spearman correlation analysis, Redundancy analysis (RDA), and Random forest (RF) analysis revealed that pH, soil organic matter (SOM), and water content (WAF) were the main drivers of bacterial community changes in the Xihe soil- covered tailings, while Sb-As concentrations and their chemical fractions were the main drivers of bacterial community structure and functional variation in the Qinglong tailings.