BackgroundAnti-endomysial antibodies (EMA) and anti-tissue transglutaminases (a-tTg) play a pivotal role in coeliac disease (CD) diagnosis. Deamidated anti-gliadin peptides (DGP) were added to the CD diagnostic workup. AimsWe aimed to compare the diagnostic accuracies of ELISA-based (a-tTg/DGP) and immunofluorescence–ELISA-based strategies (EMA/a-tTg) for CD diagnosis. MethodsFrom November 2020 to November 2022, we undertook an observational prospective study including consecutive adult patients with suspected CD. All subjects were tested for EMA, a-tTg and DGP IgA. An ROC curve was plotted to establish the best specificity cut-off of a-tTg and DGP levels, which would predict the presence of Marsh≥2 and Marsh=3. The diagnostic accuracies of a-tTg/DG and EMA/a-tTg were compared. ResultsThe study included 275 CD patients. Histology showed Marsh=1 in 9.9%, Marsh=2 in 4.5%, and Marsh=3 in 85.6.%. The best cut-off value of a-tTg for predicting Marsh≥2 was 42 U/mL, while the best cut-off for predicting atrophy was 68.4 U/mL. The best cut-off value of DGP for predicting Marsh≥2 was 56 U/mL, while the best cut-off for predicting atrophy was 78 U/mL. A-tTg/EMA showed 97% sensitivity and 100% specificity, whereas a-tTg/DGP showed 94% sensitivity and 100% specificity. ConclusionA-tTg/DGP is accurate for CD diagnosis and could reduce costs and operator-dependency of EMA. DGP, together with a-tTg, could replace EMA in CD diagnosis.