The relationship between circulating metabolites and sarcopenia-related phenotypes remains unclear. We explored the causality between circulating metabolites and sarcopenia-related phenotypes. Instrumental variables for the human metabolome were derived from the recently published GWAS, which included 690 plasma metabolites. Summary statistics for four sarcopenia phenotypes (whole-body lean mass (WBLM), usual walking pace, appendicular lean mass (ALM), and handgrip strength (HGS)) (both sexes, males and females) were obtained from relevant GWASs. We used MR to evaluate the association between circulating metabolites and sarcopenia-related phenotypes. Colocalization analysis was utilized to determine whether two associated signals were consistent with a shared causal variant rather than the confounding effect of linkage disequilibrium. Subsequently, we explored associations between modifiable risk factors and sarcopenia-related metabolites to explore which metabolites may serve as potential intervention targets through lifestyle modification. Genetically predicted plasma levels of 95 known metabolites were associated with sarcopenia-related phenotypes, and 27 metabolites were supported by robust evidence of colocalization, among which 13 metabolites had a cross-sarcopenia effect. These metabolites primarily included acyl carnitines, amino acids and their derivatives, and phospholipids. Specifically, our analyses supported causal relationships between 23, 6, and 15 metabolites and ALM, HGS, and WBLM, respectively. Seven relevant metabolites might be associated with six modifiable factors. We identified 27 metabolite biomarkers with robust causal evidence for sarcopenia-related phenotypes, highlighting 13 metabolites with a cross-sarcopenia effect, and prioritized several metabolites as the potential interventional targets of lifestyle changes. Our study provided new insight into the etiology and prevention of sarcopenia.