: BDE-47, a pervasive environmental pollutant detected in >90% of human serum samples, is increasingly linked to metabolic disorders. This study investigates the specific impact of BDE-47 exposure on the gut microbiota in prediabetic mice and evaluates the efficacy of therapeutic interventions in mitigating these effects. : To determine whether BDE-47 exposure induces diabetogenic dysbiosis in prediabetic mice and to assess whether dietary interventions, such as grape exosomes and an antioxidant cocktail, can restore a healthy microbiota composition and mitigate diabetes risk. : In this study, a prediabetic mouse model was established in 54 male SPF-grade C57BL/6J mice through a combination of high-sugar and high-fat diet feeding with streptozotocin injection. Oral glucose tolerance tests (OGTT) were conducted on day 7 and day 21 post-modeling to assess the establishment of the model. The criteria for successful model induction were defined as fasting blood glucose levels below 7.8 mmol/L and 2 h postprandial glucose levels between 7.8 and 11.1 mmol/L. Following confirmation of model success, a 3 × 3 factorial design was applied to allocate the experimental animals into groups based on two independent factors: BDE-47 exposure and exosome intervention. The BDE-47 exposure factor consisted of three dose levels-none, high-dose, and medium-dose-while the exosome intervention factor included three modalities-none, Antioxidant Nutrients Intervention, and Grape Exosomes Intervention. Fresh fecal samples were collected from mice two days prior to sacrifice. Cecal contents and segments of the small intestine were collected and transferred into 1.5 mL cryotubes. All sequences were clustered into operational taxonomic units (OTUs) based on defined similarity thresholds. To compare means across multiple groups, a two-way analysis of variance (ANOVA) was employed. The significance level was predefined at α = 0.05, and -values < 0.05 were considered statistically significant. Bar charts and line graphs were generated using GraphPad Prism version 9.0 software, while statistical analyses were performed using SPSS version 20.0 software. : The results of 16S rDNA sequencing analysis of the microbiome showed that there was no difference in the α diversity of the intestinal microbiota in each group of mice ( > 0.05), but there was a difference in the Beta diversity ( < 0.05). At the gate level, the abundances of , , , and in the medium-dose BDE-7 group were higher than those in the model control group ( < 0.05). The abundance of was lower than that of the model control group ( < 0.05). The abundances of and in the high-dose BDE-7 group were higher than those in the model control group ( < 0.05). The abundance of and was lower than that of the model control group ( < 0.05), while the abundance of in the grape exosome group was higher than that of the model control group ( < 0.05). The abundance of was lower than that of the model control group ( < 0.05), while the abundance of Firmicutes and in the antioxidant nutrient group was higher than that of the model control group ( < 0.05). However, the abundance of and was lower than that of the model control group ( < 0.05). At the genus level, the abundances of and unclassified in the high-dose BDE-7 group were higher than those in the model control group ( < 0.05). The abundance of NK4A136_group and was lower than that of the model control group ( < 0.05). The abundance of Veillonella and Helicobacter in the medium-dose BDE-7 group was higher than that in the model control group ( < 0.05), while the abundance of Lactobacillus was lower ( < 0.05). The abundance of genera such as and in the grape exosome group was higher than that in the model control group ( < 0.05). The abundance of and Bacteroides was lower than that of the model control group ( < 0.05). In the antioxidant nutrient group, the abundance of and was higher than that in the model control group ( < 0.05). However, the abundance of and UCG-002 was significantly lower than that of the model control group ( < 0.05). : BDE-47 induces diabetogenic dysbiosis in prediabetic mice, which is reversible by dietary interventions. These findings suggest that microbiota-targeted strategies may effectively mitigate the diabetes risk associated with environmental pollutant exposure. Future studies should further explore the mechanisms underlying these microbiota changes and the long-term health benefits of such interventions.