In recent years, observational studies have been conducted to investigate the potential impact of vitamins on sepsis. However, many of these studies have produced inconsistent results. Our Mendelian randomization (MR) study aims to evaluate the causality between vitamins and sepsis from a genetic perspective., Our MR study was designed following the STROBE-MR guidelines. Genetic instrumental variables for vitamins including folate, vitamin B12, B6, A (Retinol), C, D, and K were obtained from previous genome-wide association studies (GWAS) and MR studies. Five different sepsis severity levels were included in the analysis. The genetic instrumental variables were screened for potential confounders using PhenoScanner V2. MR analysis was performed using MR-egger, inverse-variance weighted multiplicative random effects (IVW-RE), inverse-variance weighted multiplicative fixed-effects (IVW-FE), and wald ratio methods to assess the relationship between vitamins and sepsis. Sensitivity analysis was performed using the MR-egger_intercept method, and the MR-PRESSO package and Cochran's Q test were used to evaluate the heterogeneity of the instrumental variables., Our MR study found no statistically significant association between vitamins and sepsis risk, regardless of the type of vitamin (P-value > 0.05). The odds ratios (ORs) for folate, vitamin B6, vitamin B12, vitamin A, vitamin D, vitamin K, and vitamin C were 1.164 (95% CI: 0.895-1.514), 0.987 (95% CI: 0.969-1.005), 0.975 (95% CI: 0.914-1.041), 0.993 (95% CI: 0.797-1.238), 0.861 (95% CI: 0.522-1.42), 0.955 (95% CI: 0.86-1.059), and 1.049 (95% CI: 0.911-1.208), respectively. Similar results were observed in subgroups of different sepsis severity levels., Our MR study found no evidence of a causal association between vitamins and sepsis risk from a genetic perspective. Further randomized controlled trials are necessary to confirm these results.