Background Smoking demonstrates divergent effects in Crohn's disease (CD) and ulcerative colitis (UC). Smoking frequency is greater in CD and deleterious to its disease course. Conversely, UC is primarily a disease of nonsmokers and ex-smokers, with reports of disease amelioration in active smoking. Aim To determine the prevalence of smoking and its effects on disease progression and surgery in a well-characterised cohort of inflammatory bowel diseases (IBD) patients. Methods Patients with smoking data of the Sydney IBD Cohort were included. Demographic, phenotypic, medical, surgical and hospitalisation data were analysed and reported on the basis of patient smoking status. Results 1203 IBD patients were identified comprising 626 CD and 557 UC with 6725 and 6672 patient-years of follow-up, respectively. CD patients were more likely to smoke than UC patients (19.2% vs. 10.2%, P < 0.001). A history of smoking in CD was associated with an increased proportional surgery rate (45.8% vs. 37.8%, P = 0.045), requirement for IBD-related hospitalisation (P = 0.009) and incidence of peripheral arthritis (29.8% vs. 22.0%, P = 0.027). Current smokers with UC demonstrated reduced corticosteroid utilisation (24.1% vs. 37.5%, P = 0.045), yet no reduction in the rates of colectomy (3.4% vs. 6.6%, P = 0.34) or hospital admission (P = 0.25) relative to nonsmokers. Ex-smokers with UC required proportionately greater immunosuppressive (36.2% vs. 26.3%, P = 0.041) and corticosteroid (43.7% vs. 34.5%, P = 0.078) therapies compared with current and never smokers. Conclusions This study confirms the detrimental effects of smoking in CD, yet failed to demonstrate substantial benefit from smoking in UC. These data should encourage all patients with IBD to quit smoking.