ObjectiveObservational studies have shown an association between COVID‐19 and epilepsy. However, causality remains unproven. This study aimed to investigate the causative effect of genetically predicted COVID‐19 phenotypes on epilepsy risk using a two‐sample Mendelian randomization (MR) analysis.MethodsWe retrieved summary‐level datasets for three COVID‐19 phenotypes (COVID‐19 susceptibility, COVID‐19 hospitalization, and COVID‐19 severity) and epilepsy from the genome‐wide association studies conducted by the COVID‐19 Host Genetics Initiative (COVID‐19 HGI) and International League Against Epilepsy (ILAE) consortium, respectively. To analyze the final results, nine MR analytic methods were utilized. The inverse‐variance weighted (IVW) method was chosen as the primary approach for data analysis to evaluate the potential causal effect. Other MR analytic methods (MR‐Egger regression, weighted median estimator, mode based‐estimator, and MR‐PRESSO) were used as a supplement to IVW to ensure the robustness of the results.ResultsThe IVW approach demonstrated no causal association between any genetically predicted COVID‐19 phenotype and the risk of epilepsy [COVID‐19 susceptibility: odds ratio (OR) = 0.99, 95% confidence interval (CI) = 0.86 ‐ 1.14, p = 0.92; COVID‐19 hospitalization: OR = 1.00, 95% CI = 0.96 ‐ 1.04, p = 0.95; COVID‐19 severity: OR = 0.99, 95% CI = 0.96 ‐ 1.01, p = 0.25]. Other MR complementary methods revealed consistent results. Additionally, no evidence for heterogeneity and horizontal pleiotropy was found.SignificanceThis MR study revealed no genetically predicted causal relationship between COVID‐19 phenotypes and epilepsy.