A case-control study of dietary factors and gastric cancer was conducted between September 1986 and March 1989 in the Barcelona metropolitan area, Spain. In all 117 cases with histologically confirmed diagnosis of gastric adenocarcinoma were matched on sex, age and possession of a telephone to 234 population controls. Of the controls 188 (80.3%) were selected by random digit telephone dialing and 46 (19.7%) by neighbourhood of residence. Information about frequency and amount of consumption of 89 alimentary items was gathered by questionnaire, and cases and controls were interviewed in their homes by trained interviewers. The gastric cancer risk decreased in proportion to vitamin C intake. In multivariate analysis adjusting for major covariables, energy and vitamin A intake, the estimated odds ratio (OR) for the upper quartile of vitamin C intake was 0.3 (95% confidence interval [CI]: 0.1-0.8). After adjustment for major covariables, calories and vitamin C intake, vitamin A did not show significant association with the gastric cancer risk. Our findings are consistent with previous case-control studies and with the hypothesis that vitamin C may inhibit the intragastric nitrosation process.