Oesophageal and gastric adenocarcinoma share an unexplained male predominance, which would be explained by the hypothesis that oestrogens are protective in this respect. We carried out a nested case-control study of hormone replacement therapy (HRT) among 299 women with oesophageal cancer, 313 with gastric cancer, and 3191 randomly selected control women, frequency matched by age and calendar year in the General Practitioners Research Database in the United Kingdom. Data were adjusted for age, calendar year, tobacco smoking, alcohol consumption, body mass index, hysterectomy, and upper gastrointestinal disorders. Among 1 619 563 person-years of follow-up, more than 50% reduced risk of gastric adenocarcinoma was found among users of HRT compared to nonusers (odds ratio (OR), 0.48, 95% confidence interval (CI) 0.29-0.79). This inverse association appeared to be stronger for gastric noncardia (OR 0.34, 95% CI 0.14-0.78) and weaker for gastric cardia tumours (OR 0.68, 95% CI 0.23-2.01). There was no association between HRT and oesophageal adenocarcinoma (OR 1.17, 95% CI 0.41-3.32).