There are inconsistent findings on the association between Helicobacter pylori anti-VacA antibodies and gastric cancer (GC) risk. The lack of optimally sensitive and specific methods to detect anti-VacA antibodies may partly be responsible for this discrepancy. The aim of this study was to compare the association between GC and the presence of anti-VacA antibodies using two different detection methods. A secondary analysis of sera collected in a hospital-based case-control study in three geographical areas of Mexico was performed. Anti-VacA antibodies were determined by a neutralization assay and an immunoblot assay in serum samples of 203 histologically confirmed GC cases and 430 age-and sex-matched controls. H. pylori IgG status was determined by immunoblotting and H. pylori CagA status data was available for this study. Unconditional logistic regression models were used to estimate the association between anti-VacA antibodies and GC by histological type (diffuse and intestinal). Anti-VacA seroprevalence was higher using neutralization compared with immunoblotting: 68.5 vs 44.3 % for cases and 60.5 vs 44.2 % for controls. A significant association between anti-VacA antibodies and diffuse GC was found using neutralization [odds ratio (OR)51.69, 95 % CI 1.08-2.66], but the association did not remain significant after adjusting for CagA status (OR 1.37,. No association between anti-VacA antibodies and GC was found when using immunoblotting. Thus, the association between anti-VacA antibodies and GC partly depends on the detection method used. These results do not support an independent role for VacA in GC risk in the presence of CagA seropositivity and strengthen the importance of CagA as a potential risk factor for GC. INTRODUCTIONLess than 3 % of subjects infected with Helicobacter pylori develop gastric cancer (GC) (Peek & Blaser, 2002). Several virulence factors produced by H. pylori are now thought to be implicated in the carcinogenesis process (Xia & Talley, 2001;Hocker & Hohenberger, 2003). Of the bacterial factors associated with GC, the most extensively studied are probably the cytotoxin CagA (Hatakeyama, 2003) and the vacuolating cytotoxin VacA (de Figueiredo Soares et al., 1998). However, other virulence factors also have activities that could induce cell transformation by affecting cell metabolism and/or causing tissue damage in general. Among these are the BabA2 adhesin, whose presence is associated with the existence of pre-neoplastic gastric lesions (Yu et al., 2002), and a urease, an enzyme that releases ammonia, which has been suggested to be the cause of an acceleration of cytokine-induced apoptosis in gastric epithelial cells (Igarashi et al., 2001). A novel apoptosis-inducing protein encoded by the H. pylori HP1118 gene has been isolated (Shibayama et al., 2003), and recently it was demonstrated that the co-chaperonin GrosES is implicated in the induction of pro-inflammatory cytokine and neoplastic changes (Lin et al., 2006). The H. pylori VacA cytotoxin is a virulence factor potentially linked...