The benefits of deep disease healing need evaluation by long-term clinical research in different populations. Confocal laser endomicroscopy (CLE) is a superior method for evaluating deep disease healing. This prospective study enrolled ulcerative colitis (UC) patients in clinical remission who underwent colonoscopy, CLE, and histological assessment. Patients were monitored for relapse by patient-reported outcomes and colonoscopy evaluation of mucosal healing. The ability of different methods of mucosal healing to predict long-term disease recurrence was assessed using Kaplan-Meier estimation and Cox proportional hazard regression. Forty-two patients in clinical remission were assessed by colonoscopy. Those with Mayo endoscopic subscores (MES)≤1 were enrolled. The 48-month recurrence rates in present healing group, assessed by CLE (colonic barrier assessment and ENHANCE index) and by histological examination (Geboes scale), were 20.0%, 26.7%, and 11.1%, respectively, and were significantly lower than absent healing group (P<0.05). Univariate Cox proportional risk regression analysis in absent of healing disease, determined by the ENHANCE index and Geboes scale, indicated an increased risk of recurrent events, with hazard ratios (HR) of 3.87 (95% CI: 1.18, 12.62) and 8.20 (95% CI: 1.06, 63.30), respectively. Multivariate Cox proportional hazard regression analysis adjusted for the extent of inflammation (E3 or not) showed a significant difference only for the ENHANCE index, with an HR of 3.53 (95% CI: 1.03, 12.10), P=0.045. Deep disease healing has a lower recurrence rate. The colonic barrier healing assessment, ENHANCE index, and histological Geboes scale have superior long-term prognostic value for UC patients.