OBJECTIVES: To compare the cost-effectiveness of once-daily tramadol extended-release (ER) and branded and generic tramadol immediate-release (IR) formulations for the treatment of chronic osteoarthritis pain from a managed care payer perspective. METHODS: A one-year model was constructed to compare the cost per percent pain reduction using tramadol formulations for treating chronic osteoarthritis pain. Prevalence, clinical efficacy, and model assumptions were based on product labels, clinical study reports, and published literature. Overall costs included: drug costs (Red Book), concomitant drug costs to treat adverse events (AEs), and resource utilization costs (office visits, emergency room visits, and inpatient hospitalizations). Effectiveness was defined as percent pain reduction, calculated as mean change from baseline in pain intensity score (ER 35.37% and IR 29.63%). Based on the literature a 30% pain reduction is considered clinically meaningful (Farrar 2001). In the cost-effectiveness analysis, a linear relationship across all costs and effectiveness ranges was assumed to extrapolate costs per clinically meaningful pain reduced (30%). Univariate sensitivity analyses were conducted to determine model inputs with the most influence on model results. RESULTS: The overall annual cost of therapy per patient was $8238 (ER), $8120 (branded IR), and $7561 (generic IR). The annual patient cost for every percentage pain reduction was lowest for ER ($232.90) followed by generic IR ($255.18) and branded IR ($274.04). The incremental cost-effectiveness ratio (ICER) for ER versus branded IR was $20.48 and the ICER for ER versus generic IR was $118.00 per percentage pain reduction. Sensitivity analyses indicated that the drug cost for ER has the most influence on the cost-effectiveness ratio. CONCLUSION: This analysis suggests the drug acquisition cost of ER may be offset by its clinical effectiveness, resulting to be a more cost-effective treatment alternative.