Background: Dye chromoendoscopy (DCE) is the recommended endoscopic modality for dysplasia detection in inflammatory bowel disease (IBD) surveillance, but randomised trials suggest comparable performance with virtual chromoendoscopy (VCE) and high-definition white light endoscopy (WLE). This multicentre UK study evaluated the association between modality and dysplasia detection in routine practice. Methods: We performed a multicentre study across six UK centres, identifying adults with ulcerative colitis (UC), colonic Crohn’s disease (CD), or primary sclerosing cholangitis associated IBD (PSC-IBD) undergoing surveillance colonoscopy between 2019-2024. Outcomes included dysplasia detection rate (DDR), neoplasia detection rate (NDR) and total lesion detection yield. Multivariable regression models and inverse probability weighting (IPW) analysis were used to evaluate detection rates taking into consideration recognised variables that associate with dysplasia risk. Findings: We included 3643 colonoscopies in 2831 patients (UC 54%, CD 22%, PSC-IBD 24%). The most frequently used modality was WLE (50%) followed by DCE (35%) and VCE (15%), with inter-centre variation in modality selection, dedicated surveillance lists, and withdrawal time. Raw DDR were higher with DCE (10%) and VCE (10%) vs. WLE (7%, p=0·012). On multivariable adjustment, modality was not independently associated with DDR. However, when variables were weighted with IPW, there was a numerically higher DDR with VCE (OR 1·60, 95% CI 1.00-2.58; p=0·051). VCE associated with higher NDR allowing the identification of more serrated and hyperplastic lesions. Longer withdrawal time, previous colorectal neoplasia, bowel preparation quality, and surveillance expertise were also associated with increased detection rates, whereas active inflammation had the opposite effect. Interpretation: In this large, multicentre study of clinical practice DCE was not superior to VCE or WLE for dysplasia detection. Identifying patients who will benefit most from surveillance and providing access to dedicated lists performed by experienced endoscopists with sufficient withdrawal time are key to increasing lesion detection in surveillance programmes.