To compare the safety and efficacy of rofecoxib* to naproxen for the treatment of juvenile rheumatoid arthritis (JRA). This was a 12-week, multicenter, randomized, double-blind, double-dummy, active comparator-controlled, non-inferiority study with a prespecified 52-week open-label active comparator-controlled extension. Children (ages 2-11 yrs) and adolescents (ages 12-17 yrs) received lower-dose (LD)-rofecoxib 0.3 mg/kg/day up to 12.5 mg/day (base study only); or higher-dose (HD)-rofecoxib (0.6 mg/kg/day up to 25 mg/day) or naproxen 15 mg/kg/day as oral suspensions. Adolescents received daily rofecoxib (LD) 12.5 (base study only) or (HD) 25 mg, or naproxen 15 mg/kg/day (maximum 1,000 mg/day) as tablets. The primary endpoint was the time-weighted average proportion of patients meeting the American College of Rheumatology Pediatric-30 (ACR Pedi 30) response criteria. A prespecified bound for the 95% confidence interval for the ratio of the percentage of ACR Pedi 30 responders was used to assess non-inferiority of treatment response between groups. Safety was assessed throughout the study. A total of 310 patients ages 2-17 years (181 (3/4) age 11) were randomized to receive LD-rofecoxib (N=109), HD-rofecoxib (N=100), or naproxen (N=101). The ACR Pedi 30 response rates following 12 weeks of treatment were 46.2%, 54.5%, and 55.1%, respectively. The relative rates of response compared to naproxen were 0.81 (95% CI 0.61, 1.07) and 0.98 (95% CI 0.76, 1.26) for LD- and HD-rofecoxib, respectively. Both rofecoxib doses were not inferior to naproxen. Patients (N=227) entering the extension received HD-rofecoxib or naproxen with efficacy maintained during the extension. All treatments were generally well tolerated throughout the study. Daily treatment of JRA patients with rofecoxib up to 12.5 or 25 mg was well tolerated, providing sustained clinical effectiveness comparable to naproxen 15 mg/kg. *On September 30, 2004, Merck & Co., Inc. announced the voluntary worldwide withdrawal of rofecoxib from the market.