Background The etiology and high share of Proximal Gastric Cancer (PGC) in China indicates that it is a distinct entity from that in other countries.PGC has a poor outcome compared with Distal Gastric Cancer (DGC),even after curative resection.Prognostic analysis may be of help in the design of clinical trials on adjuvant treatments of proximal gastric cancer.Patients and method Patient data including clinical and pathological information derived from a database established in Beijing Cancer Hospital were analyzed.Comparison between PGC and DGC was performed.Univariate and multivariate prognostic analysis on PGC following curative resection were carried out.Results Ninety hundred and thirty-eight eligible patients were enrolled in this study.Disease-specific survival of 372 PGC was superior to that of DGC,even stratified by stage.PGC had more male,larger body mass index (BMI),more advanced lesion,and more aggressive.The median number of lymph nodes harvested in PGC was 11 (1-46),lessthan DGC (19,1-71).Multivariate analysis demonstrated that lymph node metastasis ratio (LMR),Laurens classification,and Lympha Vascular Invasion (LVI)were independent prognostic factors.Proximal gastric cancer with different LMR had significant differentprognosis,even in those total lymph nodes less than 15.Conclusion LMR,LVI,and Laurens classification were independent prognostic factors for Chinese patients with proximal gastric cancer following curative resection.Clinical trials on adjuvant treatment of proximal gastric cancer should be based on prognostic stratification.