Previous studies have associated different beverage types with frailty and sarcopenia, it remains uncertain whether these associations are causal. This Mendelian randomization study aimed to investigate the causal effects of various beverage consumption on frailty and sarcopenia-related traits. Independent genetic variants strongly (P < 5E-8) associated with sweet and bitter beverages and their subtypes were used as instrumental variables. Summary-level data from recent genome-wide association studies on frailty index (FI), appendicular lean mass (ALM), and low hand grip strength (LHGS) were utilized, with false discovery rate (FDR) adjustment for multiple comparisons. Genetically predicted consumption of bitter non-alcoholic beverages (BNaBs) was associated with increased FI (β = 0.16, 95%CI: 0.04-0.28, P=0.020) and decreased ALM (β=-0.18, 95%CI: -0.31 to -0.06, P=0.019). Moreover, total bitter beverage and BNaBs consumption showed a suggestive association with an elevated risk of LHGS (P<0.05). However, the consumption of total sweet beverages, but not sugar-sweetened beverages, was associated with decreased FI and increased ALM. No causal effects on frailty and sarcopenia-related traits were observed for coffee, tea, or bitter alcoholic beverages. These findings suggest that the consumption of bitter beverages, particularly BNaBs, may be linked to the development of muscle weakness and frailty.