Reports in the literature regarding the relationship of Helicobacter pylori infection to gastric cancer are conflicting. The aim of this study was to identify the source of heterogeneity between studies. Meta-analysis of cohort or case-control studies with age- and/or sex-matched controls, providing raw data on H. pylori infection detected by serology, was used. A fully recursive literature search identified 19 qualified studies with 2491 patients and 3959 controls. Test for homogeneity found a significant difference in odds ratio between patients with early and advanced gastric cancer (6.35 vs. 2.13; P = 0.01), patients with cardiac and noncardiac gastric cancer (1.23 vs. 3.08; P = 0.003), and population- and hospital-based controls (2.11 vs. 1.49; P < 0.001). The summary odds ratio for gastric cancer in H. pylori-infected patients is 1.92 (95% confidence interval [CI], 1.32-2.78), 2.24 (95% CI, 1.15-4.4), and 1.81 (95% CI, 1.16-2.84) for all studies, cohort, and case-control studies, respectively. H. pylori-infected younger patients have a higher relative risk for gastric cancer than older patients with odds ratios decreasing from 9.29 at age < or = 29 years to 1.05 at age > or = 70 years. H. pylori infection is equally associated with the intestinal or diffuse type of gastric cancer. H. pylori infection is a risk factor for gastric cancer. The heterogeneity of reported results is caused by differences in the selection of controls, patient age, and the site and stage of gastric cancer.