To associate Helicobacter pylori-associated antibodies with clinical disease in groups of patients with duodenal ulcer, gastric adenocarcinoma, oesophageal adenocarcinoma and normal mucosa. Prospective observational sero-epidemiology study. Identification of consecutive in-patients with duodenal ulcer, gastric adenocarcinoma, oesophageal adenocarcinoma and normal mucosa. Analyses of sera for antibodies to whole H. pylori, Cag A and Vac A antigens using ELISA and Western blot. Statistical analyses. Walsgrave Hospital, Coventry, a district general hospital that serves a population of 350,000. Consecutive in-patients with an endoscopic diagnosis of duodenal ulcer (n = 31), gastric adenocarcinoma (n = 31), oesophageal adenocarcinoma (n = 40) and normal mucosa (n = 46). A profile of antibodies was constructed for each patient group and between-group comparisons were made. A logistic regression model determined the H. pylori-associated antibody that could best predict a patient's diagnosis. A discriminatory power for each antibody was calculated. Whole H. pylori, Cag A and Vac A antibodies are found more commonly in duodenal ulcer patients when compared to oesophageal adenocarcinoma (P < 0.003) and normal mucosa patients (P < 0.015). Similarly, gastric adenocarcinoma patients have antibodies to whole H. pylori, Cag A and Vac A more frequently than oesophageal adenocarcinoma (P< 0.002) and normal mucosa patients (P < 0.006). Vac A antibodies discriminate between duodenal ulcer/gastric adenocarcinoma and oesophageal adenocarcinoma/normal mucosa patients (odds ratio 5.56, log likelihood -90.06, P < 0.001) more effectively than Cag A antibodies (odds ratio 4.17, log likelihood -91.88, P < 0.001). Similar profiles of H. pylori-associated antibodies are seen in patients with duodenal ulcer and gastric adenocarcinoma, confirming that virulent H. pylori are involved in the pathogenesis of both diseases. Antibodies to Vac A could be used to identify patients at increased risk of developing H. pylori-associated disease.