BACKGROUND & AIMS: There are few validation studies of existing diagnostic criteria for irritable bowel syndrome (IBS). We conducted a validation study of the Rome and Manning criteria in secondary care. METHODS: We collected complete symptom, colonoscopy, and histology data from 1848 consecutive adult patients with gastrointestinal symptoms at 2 hospitals in Hamilton, Ontario; the subjects then underwent colonoscopy. Assessors were blinded to symptom status. Individuals with normal colonoscopy and histopathology results, and no evidence of celiac disease, were classified as having no organic gastrointestinal disease. The reference standard used to define the presence of true IBS was lower abdominal pain or discomfort in association with a change in bowel habit and no organic gastrointestinal disease. Sensitivity, specificity, and positive and negative likelihood ratios, with 95% confidence intervals, were calculated for each diagnostic criteria. RESULTS: In identifying patients with IBS, sensitivities of the criteria ranged from 61.9% (Manning) to 95.8% (Rome I), and specificities from 70.6% (Rome I) to 81.8% (Manning). Positive likelihood ratios ranged from 3.19 (Rome II) to 3.39 (Manning), and negative likelihood ratios from 0.06 (Rome I) to 0.47 (Manning). The level of agreement between diagnostic criteria was greatest for Rome I and Rome II (κ = 0.95), and lowest for Manning and Rome III (κ = 0.59). CONCLUSIONS: Existing diagnostic criteria perform modestly in distinguishing IBS from organic disease. There appears to be little difference in terms of accuracy. More accurate ways of diagnosing IBS, avoiding the need for investigation, are required.