BACKGROUND: Irritable bowel syndrome (IBS) is common, affecting 10-20% of the adult population worldwide, with many people reporting ongoing symptoms despite first-line therapies. Cognitive behavioural therapy (CBT) is recommended in guidelines for refractory IBS but there is insufficient access to CBT for IBS and uncertainty about whether benefits last in the longer term. Assessing Cognitive behavioural Therapy for IBS (ACTIB) was a large, randomised, controlled trial of two forms of CBT for patients with refractory IBS. ACTIB results showed that, at 12 months, both forms of CBT for IBS were significantly more effective than treatment as usual at reducing IBS symptom severity in adults with refractory IBS. This follow-up study aimed to evaluate 24-month clinical outcomes of participants in the ACTIB trial. METHODS: In the ACTIB three-group, randomised, controlled trial, 558 adults with refractory IBS were randomly allocated to receive either therapist-delivered telephone CBT (telephone-CBT group), web-based CBT with minimal therapist support (web-CBT group), or treatment as usual (TAU group) and were followed up for 12 months. Participants were adults with refractory IBS (clinically significant symptoms for ≥12 months despite being offered first-line therapies), recruited by letter and opportunistically from 74 general practices and three gastroenterology centres in London and the south of England (UK) between May 1, 2014, and March 31, 2016. Primary outcome measures were IBS Symptom Severity Score (IBS-SSS) and Work and Social Adjustment Scale (WSAS), assessed in the intention-to-treat (ITT) population with multiple imputation. This study was a non-prespecified naturalistic follow-up and analysis of the participants of the ACTIB trial at 24 months assessing the same outcomes as the original trial. Outcome measures were completed online by participants or a paper questionnaire was posted, or telephone follow-up undertaken. The ACTIB trial is registered with the International Standard Randomised Controlled Trial Number registry, number ISRCTN44427879. FINDINGS: 24-month follow-up of outcomes was achieved for 323 (58%) of 558 participants: 119 (64%) of 186 in the telephone-CBT group, 99 (54%) of 185 in the web-CBT group, and 105 (56%) of 187 in the TAU group. At 24 months, mean IBS-SSS was 40·5 points (95% CI 15·0 to 66·0; p=0·002) lower in the telephone-CBT group and 12·9 points (-12·9 to 38·8; p=0·33) lower in the web-CBT group than in the TAU group. The mean WSAS score was 3·1 points (1·3 to 4·9; p<0·001) lower in the telephone-CBT group and 1·9 points (0·1 to 3·7; p=0·036) lower in the web-CBT group than in the TAU group. A clinically significant IBS-SSS change (≥50 points) from baseline to 24 months was found in 84 (71%) of 119 participants in the telephone-CBT group, in 62 (63%) of 99 in the web-CBT group, and in 48 (46%) of 105 in the TAU group. In total 41 adverse events were reported between 12 to 24 months: 11 in the telephone-CBT group, 15 in the web-CBT group, and 15 in the TAU group. Of these, eight were reported as gastrointestinal related, five as psychological, and six as musculoskeletal. There were no adverse events related to treatment. INTERPRETATION: At 24-month follow-up, sustained improvements in IBS were seen in both CBT groups compared with TAU, although some previous gains were reduced compared with the 12-month outcomes. IBS-specific CBT has the potential to provide long-term improvement in IBS, achievable within a usual clinical setting. Increasing access to CBT for IBS could achieve long-term patient benefit. FUNDING: UK National Institute for Health Research.