AIM:To follow the principles of evidence based medicine toreach the integrated results of these studies.METHODS:Twenty-one papers of caes-control studies wereselected,including 11 on gastric cancer,7 on precancerouslesion of stomach and 3 on lymphoma of stomach.Metaanalysis was used to sum up the odds ratios(OR)of thesestudies.RESULTS:H.pylorl vsgastric cancer(intestinal and diffusetype):the odds ratio from the fixed effect model is 3.0016(95% Cl:2.4197-3.7234,P<0.001).H.pylori vsprecancerous lesion of stomach:a random effect model wasused to calculate the summary odds ratio and its value is2.5635(95% Cl:1.8477-3.5566,P<0.01).H.pylori vslymphoma of stomach:though the quantity of literature istoo small to make Meta analysis,the data of these 3 studiesshow that lymphoma of stomach is highly associated withH.pylori infections.CoNCLUSION:Since it had been revealed that H.pyloriinfection pre-existo in gastric carcinoma and precancerouslesions,the results of Meta analysis present a strongevldenca to support the conclusion that H.pyiori infectionis a risk factor for gastric carcinoma.