Objective To evaluate the prognostic value of blood lactate (Lac) level in sepsis patients with or without diabetes.Methods 106 patients admitted to intensive care unit (ICU) of Zhongshan Hospital Affiliated to Fudan University from April 2015 to November 2016 were enrolled. The patients with age > 18 years and the length of hospital stay > 24 hours were included. Records including blood Lac, serum creatinine (SCr), white blood cell count (WBC), platelet count (PLT), sequential organ failure assessment (SOFA) on the first day of admission; minimum oxygen index (PaO2/FiO2) in 3 days after admission; mechanical ventilation, whether there was a history of diabetes, usage of biguanides, etiology control treatment, usage of continuous renal replacement therapy (CRRT) were collected. According to the level of blood Lac patients were divided into high Lac group (Lac > 2 mmol/L) and low Lac group (Lac ≤ 2 mmol/L);based on their diabetic history, sepsis patients were divided into the diabetes group and non-diabetes group. The survival curve of each group was analyzed by Kaplan-Meier regression analysis, and the factors influencing the prognosis were analyzed by multivariate Cox regression analysis.Results There were 76 males and 30 females sepsis patients, with an average age of (68.1±14.7) years old. In the 51 patients of low Lac group, there were 7 patients who suffered from diabetes. While in the 55 patients of high Lac group, there were 12 patients who suffered from diabetes. Compared with low Lac group, high Lac group had a higher age, higher SOFA score, and a lower proportion of patients who had the treatment of etiology control (allP < 0.05). There was no significant difference of blood Lac in sepsis patients with diabetes and those without diabetes (mmol/L: 3.03±2.73 vs. 2.81±2.40,P > 0.05). Kaplan-Meier survival curve analysis showed that the 90-day survival rate in the high Lac group was significantly lower than that in the low Lac group (56.36% vs. 90.20%,χ2 = 0.697,P = 0.008). The high Lac group without diabetes had lower survival rate, and the 90-day survival rate was significantly lower than that of the low Lac group without diabetes (58.14% vs. 90.90%,χ2 = 7.152,P = 0.007); there was no significant difference in 90-day survival rate between the high Lac group and the low Lac group with diabetes (50.00% vs. 85.71%,χ2 = 0.012,P = 0.914). Multivariate Cox regression analysis showed that blood Lac was an independent risk factor for the prognosis of sepsis patients [odds ratio (OR) = 3.863, 95% confidence interval (95%CI) = 1.237-12.060,P = 0.020]. After stratification according to their diabetic history, the blood Lac was an independent risk factor for the prognosis of sepsis patients without diabetes (OR = 4.816, 95%CI = 1.407-15.824, P = 0.010), but the blood Lac had no effect on the prognosis of sepsis patients with diabetes (OR = 0.000, 95%CI =0.000-1.103,P = 0.270).Conclusions The predictive value of blood Lac on sepsis patients with or without diabetes was different. The blood Lac was related with the prognosis of sepsis patients without diabetes, while further study should be conducted for the prognostic value of blood Lac in sepsis patients with diabetes, and it's possible to increase the cut-off-point of Lac level in these patients.