Underwater endoscopic mucosal resection (UEMR) is an emerging strategy for the management of colorectal polyps. We aimed to evaluate the efficacy and safety of UEMR for clinically significant (≥ 10 mm) colorectal polyps. We performed a prospective dual-centre study of polyps ≥ 10 mm undergoing UEMR between June 2014 and March 2017. Outcomes measured comprised: (1) completeness of resection at index UEMR, (2) intraprocedural and 30-day complications, (3) rates and predictors of submucosal lift, en bloc resection, polyp/adenoma recurrence and (4) pain score. Endoscopy records were correlated with histology. 85 patients underwent UEMR of 97 polyps. Resection was endoscopically complete at index UEMR in 97.9%. The median pain score was 0 (no pain). Submucosal lift was required in 29.9% and correlated with polyp size ≥ 30 mm (p = 0.03) and clip placement (p = 0.004). En bloc resection was achieved in 45.4%, and inversely correlated with polyp size ≥ 20 mm (p < 0.001). 30-day complications (4.1%) were minor and consisted of intraprocedural bleeding (n = 2) and delayed bleeding (n = 2). 60.8% attended endoscopy post-UEMR after a median interval of 6 months, with 20.3% polyp and 13.6% adenoma recurrence. Polyp recurrence was associated with piecemeal resection (p = 0.04), recurrent polyp (p = 0.02), female sex (p = 0.01) and poor access (p = 0.005). Predictors for adenoma recurrence included female gender (p = 0.01) and difficult access (p < 0.001). Recurrence rates did not differ with polyp size, site, morphology, dysplasia status, submucosal injection, patient age, or study centre. UEMR is an effective, safe and well tolerated option for significant colorectal polyps. Piecemeal resection, recurrent polyp, female gender, and difficult access are predictors of post-UEMR polyp recurrence.