OBJECTIVE To compare the analgesic and anti-inflammatory effects of the nonsteroidal anti-inflammatory drugs (NSAID), ketoprofen (2.20 and 3.63 mg/kg of body weight) and phenylbutazone (4.40 mg/kg), in an acute equine synovitis model. METHODS 4 groups of 6 horses received NSAID or saline solution in a randomized design. METHODS 24 clinically normal mares and geldings. METHODS Left intercarpal joints were injected with sterile carrageenan to induce synovitis at the same time as IV administration of NSAID or saline solution. Clinical assessments were made and synovial fluid was withdrawn at 0, 1, 3, 6, 9, 12, 24, and 48 hours. RESULTS The eicosanoids, prostaglandin E2 (PGE2) and leukotriene B4, increased in synovial fluid after synovitis induction in all horses then returned to near baseline by 48 hours. All NSAID-treated horses had decreased PGE2, compared with saline-treated horses. This effect lasted longer in phenylbutazone-treated horses than in ketoprofen-treated horses. There were no treatment effects on leukotriene B4. In saline-treated animals, lameness, joint temperature, and synovial fluid volume, protein concentration, and nucleated cells increased 3 to 12 hours after induction, with marked reduction by 48 hours. Only phenylbutazone treatment reduced lameness, joint temperature, and synovial fluid volume. CONCLUSIONS Phenylbutazone was more effective than ketoprofen in reducing lameness, joint temperature, synovial fluid volume, and synovial fluid PGE2. Results do not support lipoxygenase inhibition by either NSAID. CONCLUSIONS This reversible model induced synovial fluid alterations similar to those observed in horses with septic arthritis. Results indicate that phenylbutazone may be more useful than ketoprofen in treating acute joint inflammation.