Objective To evaluate the diagnostic value of gastrin?17( G?17) and pepsinogen( PG) for gastric cancer. Methods A multicenter cross?sectional study of patients with continuous stomach discomfort from four centers including Changhai Hospital Affiliated to Second Military Medical University, the First Hospital Affiliated to Anhui Medical University, Qinghai Provincial People′s Hospital and the First Hospital Affiliated to Zhejiang University of Chinese Medicine from May 2014 to September 2015 was conducted. Before gastroscopy, fasting serum gatrin?17 and pepsinogen were analyzed by enzyme?linked immunosorbent assay(ELISA). The efficacy of G?17 and PG were evaluated according to endoscopic and pathological results. Results Based on the results of the pathological diagnosis, 1 122 cases were enrolled and divided into chronic atrophic gastritis group ( 548 cases ) , chronic non?atrophic gastritis group ( 370 cases), and gastric cancer group(204 cases). Serum G?17 and PGⅡ levels significantly increased(P<0?05) and PGR significantly decreased( P<0?05) in gastric cancer group compared with other groups. There was no significant difference in PGⅠlevel among three groups. The cut?off value of G?17 to diagnose gastric cancer was 7 pmol/L. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of G?17 for gastric cancer were 59?31%, 70?59%, 68?54%, 30?95% and 88?65% respectively. The cut?off value of PG Ⅰ/PG Ⅱ( PGR ) to diagnose gastric cancer was 7. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PGR for gastric cancer were 41?18%, 83?01%, 75?40%, 35?00% and 86?39% respectively. The cut?off value of PGⅡto diagnose gastric cancer was 10 μg/L. The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PGⅡfor gastric cancer were 73?53%, 53?05%, 56?77%, 25?82% and 90?02% respectively. If G?17>7 pmol/L and PGR<7 was regarded as the cut?off value of diagnosis of gastric cancer, the sensitivity, specificity, accuracy, positive predictive value and negative predictive value were 25?00%, 91?29%, 79?23%, 38?93%and 84?56%respectively. If G?17>7 pmol/L and PGⅡ>10μg/L was regarded as the cut?off value, the sensitivity, specificity, accuracy, positive predictive value and negative predictive value were 48?04%, 79?74%, 73?98%, 34?51% and 87?35% respectively. If PGR<7 and PGⅡ>10 μg/L was regarded as the cut?off value, the sensitivity, specificity, accuracy, positive predictive value and negative predictive value were 33?82%, 84?86%, 75?58%, 33?17% and 85?23% respectively. Based on logistic regression analysis of the independent variables of high serum G?17 value(>7 pmol/L), low serum PGR value(<7) and high serum PGⅡvalue(>10 μg/L), their OR value were 2?592, 2?237 and 1?864 respectively, and high serum G?17 value showed the highest risk of gastric cancer. Conclusion High serum G?17 and PGⅡ, low PGR are indicators of gastric cancer. Combination of G?17 and PGR has the best diagnostic value for gastric cacer. Gastric cancer can be screened in large scale by combining G?17 and PGR in order to improve the early diagnostic rate of gastric cancer and reduce the mortality of gastric cancer in our country.