ObjectiveTo investigate the vaginal and gut microbes changes during the carcinogenesis of cervical and the auxiliary diagnostic value. To investigate the effect of microbiome-specific metabolites butyric on cervical cancer cells. MethodsWe studied 416 vaginal 16S rRNA sequencing data and 116 gut sequencing data. Reads were processed using VSEARCH. We used Shannon index, Chao1 index, Simpson diversity index, β diversity index, Linear discriminant analysis Effect Size (LEfSe), co-abundance network and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis to explore microbiome differences between groups. We constructed random forest models based on genus and verified its discriminant effect. Finally, we used the cell counting kit-8 (CCK-8) method to detect cell proliferation capacity and flow cytometry to detect apoptosis and induction of cell cycle progression. ResultsCompared to the non-cancerous population, patients with cervical cancer had unique microbial community characteristics in both vaginal and gut ecological niches. Our predictive model based on genus in two ecological regions achieved high accuracy in the diagnosis of cervical cancer (vaginal model AUC=91.58 %; gut model AUC=99.95 %). Butyric inhibited cervical cancer cell proliferation in a concentration-dependent manner and promoted apoptosis of cancer cells. ConclusionSignificant differences were found in vaginal and gut microbes in patients with cervical cancer compared to the non-cancerous population. The prediction models constructed at the genus level in both ecological sites have good diagnostic value. Microorganisms may be involved in cervical cancer progression in a metabolite-dependent way, and targeting butyric may provide therapeutic options for cervical cancer.