BACKGROUND & AIMS: There is debate over the optimal method for colonoscopic surveillance of patients with inflammatory bowel diseases. Guidelines recommend chromoendoscopy, but the value of chromoendoscopy in high-definition colonoscopy has not been proven. Furthermore, the value of random biopsies is controversial. METHODS: We performed a prospective study of 305 patients with ulcerative colitis or Crohn's colitis referred for surveillance colonoscopy at a university hospital in Sweden, from March 2011 through April 2016. Patients randomly assigned to a group that received high-definition chromoendoscopy with indigo carmine (HD-CE; n=152), collection of 32 random biopsies, and targeted biopsies or polypectomies or to a group that received high-definition white light endoscopy (HD-WLE; n=153), collection of 32 random biopsies, and targeted biopsies or polypectomies. The primary endpoint was number of patients with dysplastic lesions. RESULTS: Dysplastic lesions were detected in 17 patients with HD-CE and 7 patients with HD-WLE (P=.032). Dysplasias in random biopsies (n=9760) were detected in 9 patients: 6 (3.9%) in the HD-CE group and 3 (2.0%) in the HD-WLE group (P=.72). Of the 9 patients with dysplasia, 3 patients (33%) had primary sclerosing cholangitis-only 18% of patients (54/305) included in the study had primary sclerosing cholangitis. The number of dysplastic lesions per 10 min of withdrawal time was 0.066 with HD-CE and 0.027 with HD-WLE (P=.056). CONCLUSIONS: In a randomized trial, we found HD-CE with collection of random biopsies to be superior to HD-WLE with random biopsies for detection of dysplasia per colonoscopy. These results support the use of chromoendoscopy for surveillance of patients with inflammatory bowel diseases. ClinicalTrials.gov no: NCT01505842.