This study aims to use Mendelian randomization (MR) analysis to explore the causal associations between gut microbiota, blood metabolites, and myasthenia gravis (MG), as well as the potential mediating mechanisms. Methods: Publicly available genome-wide association study (GWAS) summary data were utilized, encompassing 473 gut microbial taxa (n=5959), 1400 plasma metabolites, and cases (197 MG patients and 354945 controls). Bidirectional MR was employed to assess causal relationships between gut microbiota and MG, with multistep MR and mediation analyses conducted to investigate the mediating roles of metabolites. Results: A total of 23 gut microbial taxa exhibited causal associations with MG (IVW, P<0.05), while MG demonstrated reverse causal effects on the abundance of two gut microbial species. Nine plasma metabolites showed causal links with MG (IVW, P<0.05), with no reverse causal influence of MG on these metabolites. CAG-269 significantly increased the protective metabolites 1-linoleoyl-GPI (18∶2) (OR=1.120, P= 0.027) and 1-palmitoyl-GPI (16∶0) (OR=1.141, P=0.017), while decreasing the risk metabolite 3-hydroxyisobutyric acid (OR=0.902, P=0.045). The genus Jiangellaceae reduced the risk metabolite phenylacetylglutamine (OR= 0.701, P=0.042), and CAG-448 lowered the risk metabolite N-acetyl-2-aminononanoic acid (OR=0.910, P= 0.049). Multivariable MR analysis revealed that the protective effect of CAG-269 was mediated through reductions in 3-hydroxyisobutyric acid (mediation effect: 14.1% ) and increases in 1-linoleoyl-GPI (18 ∶ 2) (12.8% ) and 1-palmitoyl-GPI (16 ∶ 0) (12.9% ). Conversely, the risk-enhancing effect of Jiangellaceae was mediated by reductions in phenylacetylglutamine (14.9% ), and the risk-enhancing effect of CAG-448 was mediated by reductions in N-acetyl-2-aminononanoic acid (7.4% ). Conclusion: This MR study provides genetic evidence supporting causal associations between select gut microbial taxa, plasma metabolites, and MG.