Abstract: OBJECTIVE To investigate the dynamic trajectory of neutrophil-to-lymphocyte ratio (NLR) within 14 days after admission in patients with sepsis, and evaluate its predictive value for 180-day all-cause mortality risk. METHODS This study retrospectively enrolled 297 adult patients with sepsis from the First Affiliated Hospital of Bengbu Medical University between Jan. 1, 2023 and Jan. 1, 2024. Daily neutrophil and lymphocyte counts during the first 14 days after admission were collected to calculate NLR. Restricted cubic spline was used to analyze the relationship between NLR and mortality risk. Latent variable growth model was employed to characterize the trajectory of NLR changes over time. Multivariate COX regression was applied to identify mortality risk factors. Meanwhile, Kaplan-Meier and subgroup analyses were conducted to compare mortality risks across trajectory groups. RESULTS NLR demonstrated a nonlinear relationship with mortality risk (P<0.05). Based on the dynamic changes of NLR, patients were divided into five categories: Class 1 (high-level persistent decline group, n=74), Class 2 (low-level stable group, n=12), Class 3 (moderate-level increase group, n=123), Class 4 (high-level persistent increase group, n=19) and Class 5 (moderate-level persistent decline group, n=69). After adjustment for confounding factors, Class 4 (high-level persistent increase group) exhibited the highest in-hospital mortality risk (P<0.05). Subgroup analyses confirmed significantly increased mortality risks for Class 4 across patients of different age, sex and heart failure levels without significant interactions, indicating the generalizability of NLR trajectory indicators. CONCLUSIONS For patients with sepsis, those with high-level persistent increase of NLR have a higher long-term mortality risk. NLR trajectory change can serve as an early assessment indicator for long-term mortality risk in patients with sepsis.