Early studies suggested that two Helicobacter pylori proteins, CagA and VacA, were virulence factors. Support for that hypothesis has been undermined by geographic differences in prevalence of these antigens. To identify other possible putative virulence factors by establishing a relationship between antigens and different H. pylori diseases, two commercial available immunoblot assay kits, HelicoBlot 2.0 (Genelabs Diagnostics, Singapore) and RIDA Blot Helicobacter (R-Biopharm GmbH, Darmstadt, Germany), were used to investigate the prevalence of various specific antigen seropositivity in 80 H. pylori-infected Japanese (20 each with gastritis, duodenal ulcer, gastric ulcer, or gastric cancer). The production of interleukin-8 (IL-8) in biopsy specimens was also measured by enzyme-linked immunosorbent assay (ELISA). Both assays had 100% sensitivity; specificity was 90% for HB2.0 and 80% for RIDA-BH. With the exception of the 33-35 K antigen, there was no relationship between antigens, endoscopic diagnoses, histological findings, or mucosal IL-8 levels. The 33-35 K antigen was present in 97.5% (39 of 40) patients with gastric or duodenal ulcer compared to 70% (14 of 20) those with chronic gastritis (P < 0.006). The mean IL-8 levels in the corpus was significantly higher in those with antibody to the 33-35 K antigen compared to those without (105.4+/-22 pg/mg vs 10.2+/-8.8 pg/mg) (P=0.015). There was no relationship between other antigens including CagA and production of IL-8. In conclusion, the low-molecular-weight 33-35 K antigen may play an important role in the pathogenesis of H. pylori-related disease.