Background: Redo fundoplication may be a challenging operation with often unpredictable results. Aim of this study was to evaluate if the morphologic characteristics of the failure - as assessed by barium swallow and endoscopy - could predict the outcome of Redo fundoplication.Patients and Methods: Between 1991 and September 2008, 70 patients underwent surgical revision for failed antireflux surgery. Patients were evaluated for symptoms using a detailed questionnaire, dynamic barium swallow, endoscopy, esophageal manometry and 24-h pH, before and after surgical revision. Main symptoms were: dysphagia (78%), chest or epigastric pain (76%), heartburn (59%), gas bloat syndrome (54%). On the basis of radiologic and endoscopic findings, failures were classified as: a) too tight-a-wrap (TT) in 15 pts b) slipped or disrupted fundoplication (SF) in 40 c) telescoping (TS) in 10 d) no anatomic changes (NC) in 5. Unsatisfactory results after Redo operation were defined as: 1) symptom score >10th percentile of the preop score of 100 consecutive GERD pts who underwent primary antireflux surgery or 2) persistence of an anatomic defect.Results: Mortality was nil. Follow-up was available in all but one of the patients with a mean of 66 months (12-160). When the conversion from a Nissen to a Toupet fundoplication was performed for TT, 13/15 pts were cured (87%). SF was treated with a new fundoplication in 32 pts, with mesh addition in half of them. In 8 pts a Collis lengthening procedure was necessary. These operations were only successful in 24/40 (60%). However, 7/9 pts in this group who underwent a third operation (esophago-gastric resection or a Collis gastroplasty was required in 6) had a significant improvement in symptoms, increasing the overall good results to 77.5%. In the TS group, 6 redo fundoplications and 3 Collis gastroplasty were performed: only 3/9 pts (33%) had satisfactory results and 3 required further operations (1 pt lost to F/U). In both the SF and TS, preoperative function tests were not related to the final outcome. All the pts who underwent surgical revision for NC had persistent or recurrent symptoms (partial fundoplication in 4, gastrectomy+Roux-en-Y in 1). The difference in results in the 4 groups of patients was statistically significant (p<.01, Chi-square test).Conclusions: Refundoplication is safe. Revision for TT is recommended. SF can also be cured in the majority of cases but, often, more than one operation may be required. Revision for TS is more challenging and less rewarding, probably for a more complex anatomical and functional defect. Reoperation for symptoms persistence in spite of normal anatomical findings is not indicated.