Objective To explore the predictive value of simplified acute physiological score Ⅱ (SAPS-Ⅱ) combined with lactate clearance rates (LCR) at different moments for mortality in sepsis patients. Methods A total of 188 patients with sepsis admitted in the hospital from April 2020 to February 2023 were selected, who were evaluated using the SAPS-Ⅱ scale. Spectrophotometry was used to detect blood lactate at baseline, after 6h, 12h, 24h, and 48h, then the LCR after 6h, 12h, 24h, and 48h were calculated. The patients were divided into a survival group (n=139) and a death group (n=37) based on 28 day outcome. Logistic regression analysis was used to explore the risk factors of sepsis death, and the efficacy of SAPS-Ⅱ scores combined with LCR at different moments in predicting patient death was analyzed using receiver operating characteristic (ROC) curve. Results Twelve patients fell off, and 37 died in the remaining 176 patients, the mortality rate was 21.02%. The age, temperature, random blood glucose, blood urea nitrogen, serum creatinine, and SAPS- Ⅱ scores in the death group were significantly higher than those in the survival group (P<0.05), while platelet count and LCR at all moments were significantly lower than those in the survival group (P<0.05). The LCR of the death group continued to decrease with time. The trend of changes in the survival group were opposite, and the differences in the two groups between each two moments were statistically significant (P<0.05). The SAPS-Ⅱ scores and LCR at all moments were risk factors for patient death (P<0.05). The SAPS-Ⅱ score and LCR at all moments had predictive value for patient death, and the area under ROC curve of the combined prediction was 0.921 (95%CI 0.825 - 1.000), which was higher than the individual prediction and LCR at each moment combined with SAPS II score prediction (P<0.05). Conclusion The SAPS-Ⅱ scores and LCR at different moments are all related to death of sepsis patients, and the combined prediction of death by the above indicators is highly effective.