Immune cells are known to be associated with sepsis. However, whether these associations represent a causal relationship and whether inflammatory proteins act as mediators remain unclear. A mediation Mendelian Randomization (MR) approach was employed to assess the correlation between immune cells and sepsis, along with the mediating effects of inflammatory proteins in this relationship. Inverse variance weighting (IVW) was used as the main statistical method, with MR-Egger and weighted median serving as supplements in the preliminary MR. Sensitivity analyses were implemented using Cochrane Q test, MR-Egger intercept, Mendelian Randomization Pleiotropy RESidual Sum and Outlier and leave-one-out analysis. Subsequently, we performed replication MR, meta-analysis, Reverse MR, and linkage disequilibrium score (LDSC) regression to thoroughly verify causation. In addition, we explored whether inflammatory proteins act as mediating factors in the pathway from the immune cells to sepsis. After conducting a meta-analysis of the discovery and replication cohorts, there were 4 risk and 7 protective causal effects between genetic liability in immune cells and sepsis, with no evidence of reverse causality. Among the 92 inflammatory proteins investigated, only 2 were found to be associated with sepsis. However, inflammatory proteins did not act as mediating factors. This study elucidated the critical role of specific immune cell traits in the progression of sepsis. Our findings provide a foundation for future research into targeted immunomodulatory therapies, potentially improving patient outcomes in sepsis, and offering new insights into the complex immunological dynamics of this condition.