Introduction: Radiofrequency ablation (RFA) in combination with endoscopic resection (ER) is a method of choice for treatment of early esophageal neoplasia. Complete remission of intestinal metaplasia (CR-IM) and complete remission of dysplasia (CR-D) are commonly used as the endpoints of successful treatment. The relevance of CR-IM (in patients with macroscopically normal neo-Z-line) has recently been challenged. Methods: The aim of this prospective, single center study was to assess the long-term efficacy of RFA. Main outcome measurements were complete remission of intestinal metaplasia (CR-IM) or dysplasia (CR-D) in patients with/without a complete macroscopic eradication of Barrett's esophagus and recurrence rate of IM and dysplasia. Conover one-way analysis was used to calculate the risk factors for recurrence of intestinal metaplasia. Results: The study involved 62 consecutive patients (mean age 62, range 25-86; 55 males and 7 females) undergoing endoscopic treatment for esophageal neoplasia in our center during 2009–2013. Sixty patients were diagnosed with Barrett's esophagus related neoplasia, the remaining 2 patients had squamous neoplasia. The median follow-up was 25 months (range 3-60). In 19 patients (31%), RFA was a single treatment modality while in 43 patients (69%), RFA was combined with endoscopic resection or dissection of a visible lesion The indications for endoscopic treatment were as follows: early adenocarcinoma (EAC): 23 (37,1%), early squamous carcinoma (ESC): 2 (3,2%), high-grade intraepithelial neoplasia (HGIN): 20 (32,3%), low-grade intraepithelial neoplasia (LGIN): 17 (27,4%). A total of 117 RFA treatment sessions were performed (37 with HALO 360, 80 with HALO 90). The median number of treatment sessions per patient (ER + RFA) was 2 (1-6). CR-IM and CR-D were achieved in 69% (95% CI 61-81%) and 89% (95% CI 85-97%), respectively. In 80% of patients without CR-IM, the neo-Z-line was completely normal without macroscopically visible islands or tongues of metaplastic mucosa. During the follow-up, there were 6 recurrences at the level of neo-Z-line (out of 33 patients with BE with follow up of at least 18 months after finishing the treatment; 18%) of intestinal metaplasia. In 5 of these patients, the neo-Z-line was macroscopically normal. Low-grade dysplasia (within the Z-line) recurred in 2 patients (5%). High-grade dysplasia and/or carcinoma have not recurred. The risk factors for recurrence of intestinal metaplasia were male sex, younger age and diagnosis of cancer. We did not detect buried glands beneath the new neosquamous epithelium in any patient. Conclusion: Treatment of BE with RFA results in CR-D and CR-IM in a high proportion of patients with a low recurrence rate. A majority of patients without CR-IM or with a recurrence of IM have macroscopically normal neo-Z-line. CR-IM and a recurrence of IM might not be clinically relevant endpoints in patients with macroscopically normal neo-Z-line.