We compared the longterm efficacy and safety of 2 dosages of etodolac with that of ibuprofen in the treatment of active rheumatoid arthritis (RA). The ability of etodolac to retard, arrest, reverse, or heal joint damage due to RA was also evaluated. Patients in the early stages of RA were assigned randomly to 3 parallel groups for up to 3 years of therapy: etodolac at 150 mg bid, etodolac at 500 mg bid, and ibuprofen 600 mg qid. Concurrent disease modifying antirheumatic drugs were not permitted; established low dosage corticosteroid therapy could be continued. A total of 1446 patients was enrolled. About 50% of patients completed one year; dropout rates were comparable between groups. Both etodolac dosages provided comparable efficacy to that of ibuprofen during the first 2 months; longterm assessment showed that 1000 mg/day of etodolac produced superior improvement as assessed by patients' opinions and number of swollen joints. About 2% of patients in each group achieved remission, and radiographs showed no difference in disease progression between treatments. The incidences of adverse events were comparable, although dyspepsia and rash occurred less frequently with 300 mg/day of etodolac than with 2400 mg/day ibuprofen. A higher incidence of gastrointestinal ulcers and bleeding was seen with ibuprofen. Changes in hepatic and renal function were of minor clinical significance and were similar between the 3 groups. Both dosages of etodolac were comparable to 2400 mg/day ibuprofen in treating RA. All 3 treatment regimens were well tolerated.