AIM: To evaluate the association between infection with specific strains of Helicobacter pylori and peptic ulcer in patients referred for upper gastrointestinal endoscopy. METHODS: One thousand, six hundred and twenty-six consecutive dyspeptic patients, referred to one Endoscopy Unit in Bologna, Italy, were enrolled. For each participant, a blood sample was obtained for the measurement of distinct immunoglobulin G antibodies against H. pylori lysate and cytotoxin associated gene A (cagA). A case-control study included the whole series: patients diagnosed with duodenal (n=275) or gastric (n=71) ulcer were identified and independently compared with controls with non-ulcer dyspepsia (n=1280). RESULTS: H. pylori seroprevalence (at least one positive marker) was associated with increasing age, male sex and a diagnosis of peptic ulcer. This association was stronger with duodenal ulcer (multivariate odds ratio (OR), 5.2; 95% confidence interval (CI), 3.5-7.9) than with gastric ulcer (OR, 2.3; 95% CI, 1.2-4.4). Further analyses showed that H. pylori lysate+/cagA- subjects had a moderately increased risk of duodenal (OR, 3.2), but not gastric (OR, 1.1), ulcer. When cagA+ subjects were separately compared with seronegative patients, there was a six-fold increased risk for duodenal ulcer and a three-fold increased risk for gastric ulcer. CONCLUSIONS: A strong positive association between infection with a cagA+ H. pylori strain and the presence of peptic disease was found. The seroprevalence of anti-cagA antibodies among patients with non-ulcer dyspepsia is so high (41%) to preclude its use as a pre-endoscopic screening test.