Abstract: Objective：To explore the causal relationship between gut microbiota and tinnitus based on Mendelian randomization (MR) research method. Methods：Data on the gut microbiota in individuals from the genome wide association study (GWAS) (n=18,340) using the MiBioGen consortium, and the summary statistical data of tinnitus were obtained from published data in the UK Biobank for a two sample MR study. Using statistical model inverse variance weighting (IVW) as the main result, simple mode method, model selection method (MR-Egger), weighted median method, and weighted model were used to supplement the examination of the causal relationship between gut microbiota and tinnitus. Cochran' s Q-test and MR Egger regression were used to verify the stability and heterogeneity of the results. Results：IVW analysis showed that an increase in the abundance of class.Clostridia (OR=0.9682, 95% CI: 0.9414-0.9958, P=0.0242), genus.Ruminococcus 1 (OR=0.9594, 95% CI: 0.9284-0.9916, P=0.0141), and order. Clostridiales (OR=0.9683, 95% CI: 0.9413-0.9958, P=0.0243) might reduce the risk of tinnitus; the increased abundance of phylum. Tenericutes (OR=1.0221, 95% CI: 1.0012-1.0434，P =0.0379), class. Mollicutes (OR=1.0221, 95% CI: 1.0012-1.0434, P=0.0379), genus. Desulfovibrio (OR=1.0314, 95% CI: 1.0039-1.0596, P=0.0247), and genus. unknowngenus (OR=1.0345, 95% CI: 1.0144-1.0550, P=0.0007) might increase the risk of tinnitus. No horizontal pleiotropy or heterogeneity was found in instrumental variables. Conclusion：Class. Clostridia, genus. Ruminococcus 1, and order. Clostridiales may be potential protective bacterial groups for tinnitus, while phylum. Tenericutes, class Mollicutes, genus. Desulfovibrio, and genus. unknowngenus may be potential risk factors for tinnitus.