Numerous observational studies have elucidated a connection between leukocyte telomere length (LTL) and sepsis, yet its fundamental cause remains enigmatic. Thus, the current study’s objective is to employ a bidirectional Mendelian randomization (MR) approach to scrutinize the causality between LTL and sepsis. We selected single nucleotide polymorphisms (SNPs) associated with LTL (n = 472,174) and sepsis from a genome-wide association study (GWAS), including Sepsis (n = 486,484, ncase = 11,643), Sepsis (28 day death in critical care) (n = 431,365, ncase = 347), Sepsis (under 75) (n = 462,869, ncase = 11,568), Sepsis (28 day death) (n = 486,484, ncase = 1896), and Sepsis (critical care) (n = 431,365, ncase = 1380), as instrumental variables (IVs). The inverse variance weighted (IVW) MR method was employed as the primary approach, and various sensitivity analyses were conducted to assess the validity of this instrument and potential pleiotropy. Using the IVW method, we uncovered a potential causal relationship between genetically predicted LTL reduction and increased susceptibility to sepsis, with an odds ratio (OR) of 1.161 [95% confidence interval (CI) 1.039–1.297, p = 0.008]. However, reverse MR analysis did not indicate any impact of sepsis on LTL. Our forward MR study highlights a potential causal relationship between LTL as an exposure and increased susceptibility to sepsis. Specifically, our findings suggest that individuals with genetically determined shorter LTL may be at an increased risk of developing sepsis. This may contribute to the development of novel diagnostic and therapeutic strategies for the prevention, diagnosis, and treatment of sepsis.