Abstract: Objective: To investigate the relationship between preoperative systemic immune-inflammation index (SII) and prognosis of patients with gastric carcinoma. Methods: Patients with gastric carcinoma who underwent gastrointestinal surgery at Affiliated Hospital of Qinghai University from June 2010 to June 2015 were retrospectively examined, and their SII values were calculated; correlation was determined using Spearman’s correlation coefficient. Receiver operating characteristic (ROC) curve analysis was used to determine the best threshold value of SII as well as the relationship between SII and various clinical and pathological features. The Kaplan–Meier method was used to construct survival curves whose significance was tested using the Log-rank test, and the overall survival (OS) of the two groups of patients was analyzed. A Cox proportional hazard model was used to analyze single factors and multivariate analysis of prognostic factors and to analyze the relationship between SII and prognosis of patients with gastric carcinoma. Result: Overall, 771 patients (606 men and 165 women) were included, aged 25～81 (56.91 ± 9.546) years. According to Spearman’s correlation coefficient analysis, SII positively correlated with gastric carcinoma (r=0.399, P<0.001). The optimal intercept value of SII determined using the ROC curve was 489.52. The 771 patients with gastric carcinoma were assigned into low SII (L-SII) (SII≤489.52) and high SII (H-SII) (SII>489.52) groups. Kaplan–Meier survival curve analysis showed that the survival rate of the L-SII group was evidently high (P<0.05). The 5-year survival rates of the L-SII group in stages Ⅰ, Ⅱ, and Ⅲ were significantly higher than those of the H-SII group (81.8% vs. 24.8%, 56.5% vs. 21.2%, and 62.5% vs. 24.6%, respectively; P<0.05). Univariate analysis showed that H-SII was a risk factor for poor overall survival of patients with gastric carcinoma (HR=3.919, 95%CI: 3.08 - 4.977, P<0.001), while multivariate analysis confirmed that the level of SII is an independent factor in patients with gastric carcinoma (HR=2.707, 95%CI: 2.074 - 3.533, P<0.001). SII was found to have the most accurate predictive ability after comparing its ability with other indicators to predict the prognosis of gastric carcinoma patients using ROC curve analysis. Using SII in combination with Alb, CEA, CA125, and WBC has better predictive capability than using either alone (area under the curve=0.786). Conclusions: A high preoperative SII is an independent risk factor for worse prognosis of patients with gastric carcinoma.