Discontinuation of anti–tumor necrosis factor-α treatment (anti-TNF) (infliximab and adalimumab) in patients with inflammatory bowel disease (IBD) is associated with a high relapse risk that may be influenced by endoscopic activity at the time of stopping. We assessed the relapse rate after anti-TNF withdrawal in patients with endoscopic healing and studied predictors of relapse including the depth of endoscopic healing. This was a multicenter, prospective study in adult patients with Crohn’s disease (CD), ulcerative colitis (UC), or IBD-unclassified (IBDU), with ≥6 months of corticosteroid-free clinical remission (confirmed at baseline) and endoscopic healing (Mayo <2/SES-CD <5 without large ulcers), who discontinued anti-TNF between 2018 and 2020 in the Netherlands. We performed Kaplan-Meier and Cox regression analyses to assess the relapse rate and evaluate potential predictors: partial (Mayo 1/SES-CD 3–4) versus complete (Mayo 0/SES-CD 0–2) endoscopic healing, anti-TNF trough levels, and immunomodulator and/or mesalamine use. Among 81 patients (CD: n = 41, 51%) with a median follow-up of 2.0 years (interquartile range, 1.6–2.1), 40 patients (49%) relapsed. Relapse rates in CD and UC/IBDU patients were comparable. At 12 months, 70% versus 35% of patients with partial versus complete endoscopic healing relapsed, respectively (adjusted hazard rate aHR, 3.28; 95% confidence interval CI, 1.43–7.50). Mesalamine use was associated with fewer relapses in UC/IBDU patients (aHR, 0.08; 95% CI, 0.01–0.67). Thirty patients restarted anti-TNF, and clinical remission was regained in 73% at 3 months. The relapse risk was high after anti-TNF withdrawal in IBD patients with endoscopic healing, but remission was regained in most cases after anti-TNF reintroduction. Complete endoscopic healing and mesalamine treatment in UC/IBDU patients decreased the risk of relapse. Display omitted