Determine the prevalence and distribution of gastric intestinal metaplasia (GIM) in a large cohort of patients subjected to esophagogastroscopy (EGD). Evaluate usefulness of grading the severity of gastritis, GIM and the impact of Helicobacter pylori (HP). Define the population at risk for gastric adenocarcinoma (GC) and assess the value of surveillance.In the course of 19 years, we performed 11,600 sequential EGDs in male veterans at Brooklyn, New York. Of all patients, 47 % had EGD only one time while 53 % had EGD repeated, 11 % of these had four or more EGDs. Patients with GIM were matched with equal number of controls with no GI symptoms. All gastric biopsies were processed in one laboratory, using the standardized protocol for histological staining and for grading the severity of epithelial changes.Of all patients subjected to EGD, 354 (3.05 %) were diagnosed with GIM. Compared to controls, GIM patients were older, 80 % were over 71. Regarding ethnicity, GIM was 5.4 % more frequent in 177 African Americans than in 159 Caucasians. Distribution of GIM did not differ with respect to age or ethnicity. As many as 6 %of GIM cases were diagnosed with GC. Grading of GIM severity had a predictive value, the average grade of severity in GC was 50 % higher than in non-cancer patients with GIM. Severity of gastritis was also a useful biomarker: patients with GC had more severe gastritis. Surprisingly, HP positivity had no predictive value: HP positive patients had similar distribution of GIM as the HP negative patients. Use of proton pump inhibitors in the past was unknown.Prevalence of GC in patients with GIM was more than 200 times higher than reported in normal population. Age more than 70 years and African Americans appeared to be at higher risk. Routine EGD and histological diagnosis, with simple grading of severity of epithelial changes provides a useful predictive information. Individuals with upper GI symptoms undergoing EGD with gastric biopsy benefited from routine clinical screening for GC. Patients with higher severity of GIM should enter surveillance (Tab. 1, Fig. 10, Ref. 45).