The association between nitrate levels in public and private sources of drinking water and gastric cancer mortality in Wisconsin was investigated in a case-control study. All gastric cancer deaths of Wisconsin residents from 1982 through 1985 were compared with deaths from other causes (controls), and nitrate levels in the home drinking water of these residents were determined. Nitrate measures for public sources were obtained from historic nitrate data from municipal sources that existed in 1970. Nitrate measures for private water sources were obtained by testing the wells individually at the existing residences. Controls were matched individually to gastric cancer cases with respect to sex, year of birth, year of death, Wisconsin birth, and Wisconsin residency at the time of death. Matched-pair analyses were performed on the paired data, and the following levels of nitrate-nitrogen exposure were used as indicators of exposure: 0.5, 2.5, 5.0, and 10.0 mg/l. Matched-pair analysis was also performed for which private water supply constituted exposure. Odds ratios and 95% confidence intervals were, respectively, 0.92 (0.75, 1.12); 0.97 (0.74, 1.35); 0.86 (0.69, 1.08); 1.50 (0.12, 18.25); and 1.09 (0.82, 1.47) for exposure to private well-water sources. These results did not indicate an increased risk of gastric cancer at any level.