No population-based study has evaluated the natural course of UC over three decades in non-Caucasians. We aimed to assess the long-term natural course of Korean patients with UC in a population-based cohort. This Korean population-based, Songpa-Kangdong IBD cohort included all patients (n=1013) newly diagnosed with UC during 1986-2015. Disease outcomes and their predictors were evaluated. During the median follow-up of 105 months, the overall use of systemic corticosteroids, thiopurines and antitumour necrosis factor (anti-TNF) agents was 40.8%, 13.9% and 6.5%, respectively. Over time, the cumulative risk of commencing corticosteroids decreased, whereas that of commencing thiopurines and anti-TNF agents increased. During follow-up, 28.7% of 778 patients with proctitis or left-sided colitis at diagnosis experienced proximal disease extension. A total of 28 patients (2.8%) underwent colectomy, demonstrating cumulative risks of colectomy at 1, 5, 10, 20 and 30 years after diagnosis of 1.0%, 1.9%, 2.2%, 5.1% and 6.4%, respectively. Multivariate Cox regression analysis revealed that extensive colitis at diagnosis (HR 8.249, 95% CI 2.394 to 28.430), ever use of corticosteroids (HR 6.437, 95% CI 1.440 to 28.773) and diagnosis in the anti-TNF era (HR 0.224, 95% CI 0.057 to 0.886) were independent predictors of colectomy. The standardised mortality ratio in patients with UC was 0.725 (95% CI 0.508 to 1.004). Korean patients with UC may have a better clinical course than Western patients, as indicated by a lower colectomy rate. The overall colectomy rate has continued to decrease over the past three decades.