Background: In most circumstances, subepithelial tumors lack distinct endoscopic and ultrasonographic features. Consequently, definitive diagnosis usually requires tissue acquisition and pathologic confirmation. Establishing a tissue diagnosis is difficult because the yield of forceps biopsies is low. However, prospective data evaluating tissue sampling techniques for subepithelial lesions are currently lacking. Objective: Our purpose was to prospectively determine the diagnostic yield of endoscopic submucosal-mucosal resection (ESMR) compared with forceps biopsy for lesions limited to the submucosa (third endosonographic layer) of the GI tract. Design: A prospective head-to-head comparison was performed. Setting: The study was performed in a tertiary care hospital. Patients: Study patients were 23 adults with subepithelial lesions limited to the submucosa. Intervention: All submucosal lesions underwent forceps biopsy followed by endoscopic submucosal resection. Biopsy speciments were obtained with large-capacity “jumbo" forceps. A total of 4 double passes (8 biopsy specimens) were collected from each lesion with use of the bite-on-bite technique. Endoscopic resection was then performed with an electrosurgical snare or cap-fitted endoscopic mucosal resection device. Main Outcome Measurement: The main outcome measurement was the diagnostic yield of biopsy forceps compared with endoscopic submucosal resection. Results: Twenty-three patients with lesions limited to the submucosa were identi-fied by endoscopic ultrasonography. All lesions underwent forceps biopsy followed by ESMR. The diagnostic yield of the jumbo forceps biopsy was 4 of 23 (17%), whereas the diagnostic yield of ESMR was 20 of 23 (87%) (P = .0001, McNemar test). Conclusion: In the evaluation of subepithelial lesions limited to the submucosa, ESMR has a significantly higher diagnostic yield than jumbo forceps biopsy with use of the bite-on-bite technique. Background: In most circumstances, subepithelial tumors lack distinct endoscopic and ultrasonographic features. Establishing a tissue diagnosis is difficult because the yield of forceps biopsies is low. However, prospective data evaluating tissue sampling techniques for subepithelial lesions are currently lacking. Objective: Our purpose was to prospectively determine the diagnostic yield of endoscopic submucosal-mucosal resection (ESMR) compared with forceps biopsy for lesions limited to the submucosa (third endosonographic layer) of the GI tract. Design: A Prospective head-to-head comparison was performed. Settings: The study was performed in a tertiary care hospital. Patients: Study patients were 23 adults with subepithelial lesions limited to the submucosa. Intervention: All submucosal lesions underwent forceps biopsy followed by endoscopic submucosal resection Biopsy speciments were A total of 4 double passes (8 biopsy specimens) were collected from each lesion with use of the bite-on-bite technique. Endoscopic resection was then performed with an electrosurgical snare or cap- fitted endoscopic mucosal resection device. Main Outcome Measurement: The main outcome measurement was the diagnostic yield of biopsy forceps compared with endoscopic submucosal resection. Results: Twenty-three patients with lesions limited to the submucosa were identi-fied by endoscopic ultrasonography. All lesions underwent forceps biopsy followed by ESMR. The diagnostic yield of the jumbo forceps biopsy was 4 of 23 (17%), while the diagnostic yield of ESMR was 20 of 23 (87%) (P = .0001, McNemar test). Conclusion: In the evaluation of subepithelial lesions limited to the submucosa, ESMR has a significantly higher diagnostic yield than jumbo forceps biopsy with use of the bite-on-bite technique.