Previous studies demonstrated a controversial relationship between sarcopenia (SP) and osteoarthritis (OA) and their genetic causality is unclear. Thus, we conducted a Mendelian randomization (MR) analysis to evaluate the possible causal association between sarcopenia-related traits (appendicular lean mass (ALM), grip strength, usual walking pace) and OA. We used pooled genetic data from the UK Biobank for ALM(n = 450,243), left-hand grip strength (n = 461,026), right-hand grip strength (n = 461,089) and usual walking pace (n = 459,915). Moreover, summary statistics for OA were obtained from the latest study conducted by the Genetics of Osteoarthritis Consortium, including all OA (n = 826,690), hand OA (n = 303,7782), hip OA (n = 353,388) and knee OA (n = 396,054). The primary method for estimating causal effects was the inverse-variance weighted (IVW) method, with the utilizing of false discovery rate adjusted values ( ). Additional MR methods such as MR-Egger regression, MR pleiotropy residual sum and outlier (MR-PRESSO), weighted median were employed as supplementary analyses. We discovered ALM (odds ratio (OR) = 1.103, 95% confidence interval (CI) = 1.052-1.156, = 2.87E-04), hand grip strength (left, IVW OR = 0.823, 95% CI = 0.712 to 0.952, = 0.020; right, OR = 0.826, 95% CI = 0.718 to 0.950, = 0.020), and usual walking pace (OR = 0.339, 95% CI = 0.204 to 0.564, = 2.38E-04) were causally associated with OA risk. In the reverse MR analysis, we identified a causal effect of OA on ALM (β = -0.258, 95% CI = -0.369 to 0.146, = 0.6.07E-06), grip strength (left, β = -0.064, 95% CI = -0.104 to 0.024, = 0.002; right, β = -0.055, 95% CI = -0.095 to 0.014, = 0.008), and usual walking pace (β = -0.104, 95% CI = -0.147 to 0.061, = 1.61E-05). This present study suggests an obvious causality of SP on OA, with condition exhibiting site-specific effects, while evidence was also provided for the causal effect of OA on SP.