OBJECTIVE To investigate the value of the difference between peripheral arterial and venous blood gas analysis for the prognosis of patients with septic shock after resuscitation. METHODS Patients with septic shock aged 18 to 80 years admitted to intensive care unit (ICU) of Affiliated Hospital of Integrated Traditional Chinese and Western Medicine, Nanjing University of Chinese Medicine from May 2016 to December 2017 were enrolled. The peripheral arterial blood and peripheral venous blood gas analysis were measured simultaneously after the early 6 hours resuscitation, including pH, partial pressure of oxygen (PO2), partial pressure of carbon dioxide (PCO2), base excess (BE), bicarbonate (HCO3-) and lactate (Lac) level, and the difference values between peripheral arterial and venous blood were calculated. According to the 28-day survival, the patients were divided into survival group and death group. Multiple Logistic regression analysis was used to analyze the risk factors of death, and the receiver operating characteristic curve (ROC) was used to analyze the prognostic value of blood gas analysis parameters for prognosis. RESULTS A total of 65 patients with septic shock resuscitation were enrolled in the study, 35 survived while 30 died during the 28-day period. (1) There was no significant difference in gender, age, and mean arterial pressure (MAP), central venous pressure (CVP), central venous oxygen saturation (ScvO2) and norepinephrine (NE) dose between the two groups. (2) The arterial and venous Lac, the difference of Lac (ΔLac) and PCO2 (ΔPCO2) between arterial and venous blood in death group were significantly higher than those in survival group [arterial Lac (mmol/L): 7.40±3.10 vs. 4.82±2.91, venous Lac (mmol/L): 9.17±3.27 vs. 5.81±3.29, ΔLac (mmol/L): 1.77±0.54 vs. 0.99±0.60, ΔPCO2 (mmHg, 1 mmHg = 0.133 kPa): 9.64±5.08 vs. 6.70±3.71, all P < 0.01], and there was no significant difference in the other arterial and venous blood gas analysis index and its corresponding differential difference between two groups. (3) Multiple Logistic regression analysis showed that ΔPCO2 [β = 0.247, odd ratio (OR) = 1.280, 95% confidential interval (95%CI) = 1.057-1.550, P = 0.011], and ΔLac (β = 2.696, OR = 14.820, 95%CI = 2.916-75.324, P = 0.001) were the independent risk factors for the prognosis of septic shock. (4) It was shown by ROC curve analysis that arterial blood Lac, ΔLac and ΔPCO2 had predictive value on prognosis of septic shock, the area under ROC curve (AUC) was 0.792, 0.857, 0.680, respectively (all P < 0.05). When the best cut-off value of arterial Lac was 4.00 mmol/L, the sensitivity was 100%, and the specificity was 62.86% for predictor of death in 28-day; when the best cut-off value of ΔLac was 1.25 mmol/L, the sensitivity was 93.33%, and the specificity was 68.57% for predictor of death in 28-day; when the best cut-off value of ΔPCO2 was 4.35 mmHg, the sensitivity was 83.33%, and the specificity was 37.14% for predictor of death in 28-day. CONCLUSIONS Compared to other parameters, the difference between peripheral arterial and venous blood gas analysis, ΔPCO2 and ΔLac had the best correlation with the prognosis of septic shock. The ΔPCO2 and ΔLac are the independent prognostic predictors for 28-day survival.