OBJECTIVES: To estimate incremental cost-utility ratios for lumiracoxib relative to celecoxib for treating osteoarthritis (OA) in Canada. Secondary comparators including six common treatment algorithms (non-steroidal anti-inflammatory drugs with and without proton pump inhibitors) were also evaluated. METHODS: An existing Markov model with 3 month cycle lengths and 5 year time horizon was adapted for Canada. Analyses were performed from the third-party perspective of the Ontario Ministry of Health. Treatments were assumed to be equally efficacious in treating symptoms of OA. Data on differences in rates of gastrointestinal (GI) and other (renal, skin, hepatic) adverse events were obtained from published randomized trials. Quality-adjusted life years (QALYs) were calculated separately for subgroups defined a priori for age, sex, aspirin use, and history of GI bleed. Common treatment pathways were elicited from clinical experts. Costs of hospitalizations, laboratory tests, professional fees, and medications were obtained from published sources. Societal utility scores were obtained from a standard gamble study conducted in Canadians. Costs (2006 CDN dollars) and outcomes were discounted at 5%. Incremental cost-effectiveness ratios (ICERs) were estimated relative to lumiracoxib. Sensitivity analyses were conducted for all input parameters to identify influential inputs. RESULTS: Lumiracoxib was more effective and less costly (i.e. dominated) in all subgroups when compared to celecoxib. Compared to celecoxib, lumiracoxib was predicted to reduce clinical and complicated events in non-ASA patients by 10 and 55% respectively. ICERs ranged from -$11,253/QALY to -$187,203/QALY for average risk patients and became more favorable over the cohort's life-time. Results were most sensitive to the utility of arthritis and adverse event rates but the interpretation was robust. Compared to the majority of secondary treatment algorithms, lumiracoxib also had an attractive cost-effectiveness profile. CONCLUSION: From an economic perspective, lumiracoxib is an attractive treatment choice for Canadian OA patients and is more cost-effective than celecoxib.