Helicobacter pylori causes gastric adenocarcinoma; whether treatment of H. pylori infection prevents this cancer remains unknown. In a randomized, double-blind, placebo-controlled trial of H. pylori eradication, we determined whether treatment for H. pylori decreases gastric cancer risk, using preneoplastic conditions as surrogate markers. A total of 248 healthy volunteers (age >40 years) randomly received H. pylori treatment (omeprazole, amoxicillin, clarythromycin; n = 122) or matched placebo (n = 126) for 1 week. Endoscopy was performed at baseline and at 6 weeks and 1 year. Seven biopsies from each endoscopy were reviewed by two pathologists using the revised Sydney classification. Outcome measures were both a consensus "worst biopsy" diagnosis and a weighted index score that incorporated degrees of severity of preneoplasia from all biopsies. We compared change in these outcomes over time between the two treatment groups. H. pylori cure rates for compliant subjects in the treatment arm were 79.2% and 75.7% at 6 weeks and 1 year, respectively. No statistically significant change in the worst biopsy diagnosis was observed from 6 weeks to 1 year between placebo and treated subjects (for improvement/worsening, placebo, 19.4%/10.5%; treatment, 22.5%/8.3%; P = 0.74). Change in index score was favorably greater in treatment compared with placebo subjects (intention-to-treat analysis, P = 0.03); this finding was particularly evident in the antrum. H. pylori eradication gave more favorable gastric histopathologies over 1 year than no treatment. Such incomplete regression suggests but does not prove that eradication of H. pylori decreases cancer risk.