Background: Barrett's epithelium is a risk factor for esophageal cancer. In this study we attempted to reverse Barrett's epithelium in an acid-reduced environment by using multipolar electrocoagulation. Methods: All patients had specialized columnar epithelium of at least 2 cm in length. Patients with dysplasia on successive screening examinations were excluded. Esophageal motility and 24-hour pH studies were performed before therapy and at 9 and 18 weeks. All patients received lansoprazole 30 mg twice a day regardless of pH study results. Multipolar electrocoagulation was applied to one side of the esophagus (2 to 3 cm per session) with the contralateral side serving as control. Biopsies were obtained from each 2 cm length at each endoscopy and reviewed by a single, blinded pathologist. At the 9-week evaluation of the treatment side, all patients elected to undergo multipolar electrocoagulation therapy for the control side. Results: Twenty-seven patients completed the study: 21 men and 6 women, ages 33 to 81 years, length of specialized columnar epithelium 2 to 10 cm. Endoscopy at the 18-week follow-up showed normal mucosa (n = 16), residual small (< 3 mm) islands or tongue-shaped extensions of apparent specialized columnar epithelium (n = 7), untreated distal strips (< 5 mm) of apparent specialized columnar epithelium at the gastroesophageal junction (n = 3), and persistent specialized columnar epithelium (n = 1). Histologically, only 5 of 27 patients had residual specialized columnar epithelium at week 18. Of 16 patients with a normal endoscopic appearance, one still had specialized columnar epithelium on biopsy. While on lansoprazole, pH studies were normal in 15 of 26 patients at 9 weeks. Of 22 patients with histologic evidence of specialized columnar epithelium reversal, 10 had persistently abnormal 24-hour pH studies (pH < 4, 6.1% to 33.7% of total time). Four of five patients with residual specialized columnar epithelium on biopsy had persistent reflux. Side effects of multipolar electrocoagulation were transient, and treatment did not adversely alter motility patterns. Conclusions: Multipolar electrocoagulation therapy can induce re-epithelialization with normal squamous mucosa in most patients with specialized columnar epithelium and does not adversely affect esophageal motility. Adequate acid suppression is not mandatory for therapy to be effective in the short term. (Gastrointest Endosc 1999;49:547-53.)