A hospital-based case-control study of gastric cancer precursor lesions was conducted in a high-risk black population in southern Louisiana. Ninety-three subjects with biopsy-proved advanced chronic atrophic gastritis were compared to two control series: a gastroscopy clinic series and a general hospital-admission series. Dietary case-control differences indicated a protective effect associated with fruit and vegetable intake and with dietary vitamin C and a risk elevation associated with milk consumption. The protective effect associated with consumption of fruits, vegetables, and vitamin C is consistent with findings for gastric cancer and with the etiologic hypothesis of intragastric nitrosation. A twofold increased risk was associated with cigarette smoking. Gastric juice pH, NO3-, and NO2- were determined for subjects undergoing gastroscopy, and comparisons were made between this high-risk U.S. group and a Colombian population with a much greater magnitude of risk; the latter had higher NO3- and NO2- levels. An increase in pH was associated with increasing severity of gastric lesions. Levels of pH and NO2- concentration were significantly correlated (P less than .0005); however, in Louisiana the large difference in NO2- concentration associated with pH elevation is not associated with histopathologic severity. Divergent trends with severity of lesions for NO3- concentration were seen in the two populations.