A large number of case-control studies have fairly consistently observed reduced risk of gastric cancer in different countries. Prospective studies are few, and their findings are less impressive. The Netherlands Cohort Study, which was based on a validated dietary survey with 120,000 residents, recently reported a relative risk of 0.8 (95% CI 0.5-1.3) for the highest versus lowest quintile of vegetable and fruit consumption (Am J Epidemiol 1998; 148: 842-53).