AIM: Some endoscopic features of duodenal mucosa are marker of mucosal injury, the most common cause being celiac disease (CD). The aim of this study was to prospectively assess the diagnostic value of the endoscopic markers for the diagnosis of CD in the adult population undergoing routine upper endoscopy. METHODS: This was a prospective multicenter study conducted at 37 Italian endoscopic centers. A total of 509 consecutive patients submitted to routine upper endoscopy who presented one or more of following endoscopic markers were included: 1) mucosal mosaic pattern in the bulb and/or descending duodenum (DD); 2) nodularity in the bulb and/or DD; 3) scalloping of Kerkring's folds; 4) reduction in the number or absence of folds in the DD. 4 biopsies samples were taken from descending duodenum. In patients with histological findings consistent with CD, according to Oberhuber classification, sierologic test (EMA, tTGA) were performed for confirm the diagnosis. RESULTS: At endoscopy, 249 patients showed an isolated marker; 260 subjects showed a coexistence of more than one marker; 369 patients (72.5%) presented mucosal lesions at histological examination and in 347 of these patients the diagnosis of CD was confirmed by serologic markers (94.0%). For 10 patients the diagnosis remained uncertain because of negative sierology and exclusion of other other cause of mucosal lesions. The diagnosis of CD was made in 61.3% patients who showed the mosaic pattern, in 65.7% of patients with nodular mucosa, in 64.4% of patients with scalloping of folds, in 40.2% of patients with reduction of folds, and in 61.5% of patients with loss of folds and in 83.6% of patients who showed the coexistence of more than one marker. The endoscopic markers overall had a PPV of 68% for the diagnosis of CD; the markers that singularly have demonstrated a higher correlation with CD are: mosaic mucosa of DD (PPV 65.0%), nodular mucosa of the bulb and DD (PPV 75.5%), and scalloping of folds (PPV 64.4%). CONCLUSION: The study confirms the important role of endoscopy in the diagnostic process of CD not only for the bioptic sampling in patients with clinical suspicion of CD, but especially for the opportunity to evaluate alterations of the duodenal mucosa suggestive of CD in the general population and, consequently, to identify those patients who should undergo a duodenal biopsy.