Objective To evaluate the usefulness of a commercial serological enzyme immunoassay (EIA) for detecting salivary immunoglobulin (Ig) G to Helicobacter pylori. Design Prospective, multicentre study. Setting The study was conducted at 11 gastroenterology hospital centres throughout Italy. Participants and interventions Two hundred and thirteen dyspeptic patients underwent gastroscopy with antral biopsies. At each centre, two of the following three tests for H. pylori diagnosis were performed: urease quick test, histology, and 13C‐urea breath test. Samples of unstimulated saliva and venous blood were collected from each patient. Salivary and serum H. pylori IgG were determined with the EIA Helori‐test IgG (Eurospital, Trieste, Italy). Main outcome measures Sensitivity, specificity, positive and negative predictive values, and accuracy of the salivary and serum EIAs. Results By using a receiver operating characteristic curve, salivary EIA yielded 81% (95% confidence interval 73‐87%) sensitivity, 73% (62‐83%) specificity, 84% (76‐90%) positive predictive value, 69% (58‐79%) negative predictive value, and 78% (72‐84%) accuracy. At the cut‐off suggested by the manufacturer, serum EIA had 90% (84‐95%) sensitivity, 78% (67‐86%) specificity, 88% (81‐93%) positive predictive value, 82% (71‐90%) negative predictive value, and 86% (81‐91%) accuracy. Conclusion In this large multicentre study, detection of salivary H. pylori IgG with a commercial serological EIA was a fairly accurate diagnostic method. Data confirm that saliva testing does not compare favourably with serology in the assessment of H. pylori status.