Background It has been suggested that GastroPanel might be a useful tool for the diagnosis of chronic atrophic gastritis (CAG) measuring four biomarkers in blood: basal gastrin-17 (G17), pepsinogen I and II (PGI and PGII), and Helicobacter pylori antibodies. Aim To determine the accuracy of GastroPanel for the diagnosis of CAG. Methods This was a prospective, blinded, multicenter study that included dyspeptic patients. G17, PGI, and PGII were determined by enzyme immunoassays. Three antrum and two corpus biopsies were obtained for standard histological analysis and rapid urease test. Biopsies were analyzed by a single blinded expert pathologist. Results Ninety-one patients were included (77% women, mean age 44 years, 51% H. pylori positive, 17% with CAG). G17 was reduced in patients with antrum CAG (5.4 vs. 13.4 pmol/l; P<0.01) and increased in patients with corpus CAG (11 vs. 24 pmol/l; P<0.05), but its accuracy was only acceptable in the case of corpus localization [ area under the receiver operating characteristic curve (AUC), 74%]; PGII difference was almost statistically significant only when testing for corpus atrophy (33 vs. 21 g/l; P=0.05; AUC=72%). The PGI and PGI/PGII ratio showed no significant differences (AUCs were all unacceptably low). Helicobacter pylori antibody levels were higher in H. pylori-infected patients (251 vs. 109 EIU, P=0.01; AUC=70). The accuracy of GastroPanel for the diagnosis of CAG was as follows: sensitivity 50%; specificity 80%; positive 25% and negative 92% predictive values; and positive 2.4 and negative 0.6 likelihood ratios. Conclusion GastroPanel is not accurate enough for the diagnosis of CAG; thus, its systematic use in clinical practice cannot be recommended. (C) 2014 Wolters Kluwer Health vertical bar Lippincott Williams & Wilkins.