Previous studies have shown that osteoarthritis (OA) and sarcopenia (SP) are closely related to each other, but the causal relationships between them have not been established. The aim of this study was to investigate the causal associations between OA and SP via a bi-directional Mendelian randomization (MR) approach.A bi-directional two-sample MR was adopted to research the causal relationship between SP and OA. The instrumental variables for SP and four types of OA: KOA, HOA, total knee replacement (TKR) and total hip replacement (THR) were derived from published large genome-wide association studies (GWAS). The inverse variance weighted (IVW), MR-Egger and weighted median estimator (WME) methods were used to estimate bi-directional causal effects.Low grip strength (GS) did not have a causal effect on four types of OA (KOA: OR = 1.205, 95% CI 0.837-1.734, p = 0.316; HOA: OR = 1.090, 95% CI 0.924-1.609, p = 0.307; TKR: OR = 1.190, 95% CI 1.084-1.307, p = 0.058; THR: OR = 1.035, 95% CI 0.792-1.353, p = 0.798), while appendicular lean mass (ALM) had a causal effect on four types of OA (KOA: OR = 1.104, 95% CI 1.041-1.171, p = 0.001; HOA: OR = 1.151, 95% CI 1.071-1.237, p < 0.001; TKR: OR = 1.114, 95% CI 1.007-1.232, p < 0.001; THR: OR = 1.203, 95% CI 1.099-1.316, p < 0.001). In the reverse direction, KOA or HOA did not have a significant causal effect on both GS and ALM (KOA-GS: OR = 1.077, 95% CI 0.886-1.309, p = 0.458; KOA-ALM: Beta = 0.004, p = 0.892; HOA-GS: OR = 1.038, 95% CI 0.981-1.099, p = 0.209; HOA-ALM: Beta = - 0.017, p = 0.196; TKR-GS: OR = 0.999, 95% CI 0.739-1.351, p = 0.997; TKR-ALM: Beta = 0.018, p = 0.501; THR-GS: OR = 1.037, 95% CI 0.978-1.101, p = 0.222; THR-ALM: Beta = - 0.023, p = 0.081).The present study suggests that SP may have a causal effect on OA through changes in muscle composition rather than muscle strength, while little evidence was provided for the causal effect of OA on SP.© 2023. BioMed Central Ltd., part of Springer Nature.