Introduction Azathioprine (AZA) is an established treatment for ulcerative colitis (UC). However, controversy exists regarding its efficacy in inducing and maintaining clinical remission, particularly with the advent of biologics which, unlike AZA, have been tested in large, rigorously designed randomised controlled trials. We studied the effectiveness of AZA as second-line therapy after failure of 5-aminosalicylates (5-ASAs) in a large cohort of UC patients, with particular emphasis on whether its earlier use alters the natural history of the disease course. Methods All UC patients treated with AZA at our centre were identified from a prospective electronic database. We excluded individuals who had received either infliximab or ciclosporin as a bridge to AZA. The following demographic data were collected: gender, age at diagnosis, age at AZA commencement, concomitant therapy at AZA commencement, and duration of disease prior to AZA commencement. We assessed response to therapy at 4 months and remission at last point of follow-up, using physicians’ global assessment, need for hospitalisation, escalation of therapy to a biologic, or colectomy, and serious adverse events (including infections and malignancies). We examined whether earlier AZA use (within 12 months of diagnosis) reduced need for hospitalisation, biologic therapy, or colectomy. Results In total, 255 patients were included (55% male, mean age at diagnosis 36.4 years). Mean age at commencing AZA was 42.3 years. Mean disease duration prior to AZA commencement was 70 months. Concomitant therapy at AZA commencement was oral 5-ASAs in 87%, topical 5-ASAs in 22%, and oral prednisolone in 77%. At 4 months, 207 (81%) of 255 patients were still on AZA (46 had discontinued due to adverse events and 2 due to non-response), and 163 (64%) had responded to therapy. There were 165 (65%) patients still receiving AZA at last point of follow-up, of whom 153 (60%) were in remission (mean duration of therapy 64.5 months). 26 patients required admission to hospital for an exacerbation during AZA treatment, 20 patients ultimately required biologic therapy, and 21 underwent colectomy. Among 90 patients receiving AZA within 12 months of diagnosis, 21 (23%) patients experienced one of these three endpoints, compared with 29 (19%) of 154 who commenced AZA >12 months after diagnosis (p = 0.40). Serious adverse events included 6 cases of pancreatitis, 6 cases of cancer (3 non-melanoma skin cancers) and 1 case of neutropenic sepsis presenting within 1 month of AZA commencement. Conclusion AZA is a safe and effective therapy for UC patients who fail 5-ASAs, and should continue to be used prior to instituting biologic therapies. However, earlier use does not seem to alter the natural history of the disease. Disclosure of Interest None Declared.