Barrett’s mucosa (BM) conveys an increased risk of esophageal cancer. Histological regression of BM after successful anti-reflux surgery is debated. Aim of the study was to investigate whether anti-reflux surgery may result in phenotypic and genotypic changes of short (<3cm) and long (>3 cm) BM segments. Thirty-five patiens [28M:7F, median age 52yrs (IQR 40-59)] with histologically proved BM underwent anti-reflux surgery between 1994 and 2003. Pre/post-op studies included: symptom assessment, manometry, pH/bile 24-hr monitoring and endoscopy (Seattle protocol biopsies). Histology sections were stained with H&E and blindly assessed by an expert pathologist. Intestinal metaplasia (IM) was semi-quantitatively scored (in each biopsy set: >or<30%) and classified as complete or incomplete based on HID histochemestry (complete <40%, incomplete ≥40% sulphomucin expression). The Cdx-2 expression was assessed by immunohistochemistry. BM regression was defined as a decrease in the IM score; BM disappearance as the absence of intestinalized glandular mucosa in the follow-up biopsies. Median post-op follow-up was 28 months (IQR:17-49), with >1 yr follow-up in all patients. Eleven patients (31%) had short BM [incomplete/complete IM: 1/10, non-invasive neoplasia (NiN): 0/11] and 24/35 (69%) had long BM (incomplete/complete IM: 10/14, low-grade NiN: 2/24). Symptom score, acid/bile exposure in the distal esophagus were significantly reduced after surgery (p<0.001); 4 (11%) patients still had abnormal acid reflux and were considered as failures. BM disappeared in 6 short BM patients (17%), with complete loss of Cdx-2 expression. Regression was observed in 11 (31%) patients (short vs. long BM: p=ns). No changes were detected in the remaining 18 patients. Among the 29 patients who maintained BM, IM significantly shifted from incomplete to complete (p=0.015). Low-grade NiN was no longer demonstrated in 1 of 2 pts at 25 months follow-up (the other one persisted at 15-months). BM disappearance correlated with 1) short BM and complete IM on univariate analysis and 2) presence of short BM on logistic regression analysis. These data confirm that BM is reversible. Effective antireflux surgery induces regression of BM; the interruption of the damaging stimuli (acid and bile reflux) leads to the disappearance of IM in subjects with short segment BM and prevalence of complete IM