Abstract: Objective To investigate the causal relationship between gut microbiota and obstructive sleep apnea syndrome (OSA) using a two-sample bidirectional Mendelian randomization (MR) approach. Methods Eligible single nucleotide polymorphisms (SNPs) were selected from genome-wide association study (GWAS) databases as instrumental variables. A bidirectional MR analysis was conducted to evaluate the causal effects between gut microbiota and OSA. Various statistical methods, including the inverse variance weighted (IVW) method, MR-Egger regression, the weighted model method, and the weighted median method, were employed for association assessment. The MR pleiotropy residual sum and outlier (MR-PRESSO) test, along with Cochran's Q test and the leave-one-out cross-validation method, were used to assess heterogeneity and pleiotropy. Results According to the IVW method analysis, an increased abundance of the genus Faecalibacterium (sp002397985) (OR=0.847, 95%CI, 0.719 to 0.997, P=0.046) was associated with a reduced risk of OSA. Conversely, increased abundances of the genera Bacteroides (OR=1.075, 95%CI, 1.016 to 1.138, P=0.012), Haemophilus (sp001679485) (OR=1.106, 95%CI, 1.016 to 1.203, P=0.021), Streptococcus (OR=1.168, 95%CI, 1.036 to 1.316, P=0.011), and Blautia (sp002159835) (OR=1.169, 95%CI, 1.035 to 1.319, P=0.012) were associated with an elevated risk of OSA. The reverse MR analysis revealed no significant association between the risk of OSA and the abundance of gut microbiota. The results of Cochran's Q test, MR-Egger test, and MR-PRESSO test indicated no heterogeneity or horizontal pleiotropy (P>0.05). Conclusion Causal relationships exist between the five genera (Faecalibacterium, Bacteroides, Haemophilus, Streptococcus, and Blautia) and the risk of OSA.