Gastric Cancer is the fifth most prevalent malignancy and the third leading cause of cancer death worldwide, of which more than half occurred in East Asia. Although previous genomic studies demonstrated the potential for molecular classifications of gastric cancer for personalized medicine, e.g. microsatellite Instability (MSI)-high subtype of patients showing resistant to chemotherapy, a large proportion of gastric cancer patients need further molecular investigation to develop tailored therapeutic regimen. We characterized the proteogenomic landscapes of 186 gastric cancer patients (including all Lauren classifications) by integrative analysis of genomic, transcriptomic and proteomic data. Fresh-frozen treatment-naive gastric cancer surgical samples were collected from the biobank of Peking University Cancer Hospital (2010.01-2013.06). Tumor cellularity was examined by H&E staining, and samples with tumor cellularity >=60% were selected for exome sequencing, RNA-seq and reproducible proteomic quantification, by pressure cycling technology (PCT) assisted sample preparation followed by SWATH‐MS. The mass spectrometry data were processed for consensus clustering analysis, which defined four proteomic subtypes in our cohort. We find that this molecular classification is significantly correlated with patient outcome and presents distinct molecular characteristics between subtypes. By combining genomic and proteomic features, our analysis elucidates the molecular details of gastric cancer, and identifies novel prognosis markers and therapeutic targets. Correspondence to Jianmin Wu (wujm@bjmu.edu.cn), Ruedi Aebersold (aebersold@imsb.biol.ethz.ch) and Jiafu Ji (jijiafu@hsc.pku.edu.cn).