Telomeres are considered to be a physiological marker of aging. Elucidating relationship between telomere length and sepsis is an essential step towards understanding the biological processes involved in sepsis and its salvation. Mendelian randomization studies based on SNPs have given us new insights into genetic susceptibility to disease.To explore the causal pathway from telomere length to occurrence and 28-day mortality of sepsis.Leveraging genetic information resource of UK Biobank, we captured three groups of large-scale GWAS data: leukocyte telomere length (LTL), sepsis and all-cause death of 28-day. Study design consisted of three parts: forward analysis, reverse analysis and one-way analysis. Genetic instrumental variables were selected for different analyses under the premise that three MR core assumptions were satisfied. Causality was determined by means of IVW.In forward analysis, we did not observe a significant causal pathway from sepsis to LTL under IVW model: β (SE) was -0.0051 (0.0075) with a p-value of 0.499. In reverse analysis, based on the IVW model, the OR (95% CI) was 0.89 (0.80-0.99) and the p-values was 0.043; based on the results of leave out method and single SNP analysis, we obtained seven key SNPs. There were results of IVW model in the one-way analysis: β (SE) was -0.0287(0.1261).Short LTL increases susceptibility to sepsis, but sepsis does not shorten telomere length. LTL does not affect sepsis 28-day all-cause mortality and does not serve as a causal intermediate in gene regulation during the progression of sepsis to 28-day death.