Objective-To determine the relationship between serum cortisol concentration and pain severity as measured by force platform gait analysis in dogs with experimentally induced synovitis of the stifle joint. Animals-10 healthy hound-type dogs. Procedures-Dogs underwent 2 study phases. In the first phase, serum cortisol concentration, systolic arterial blood pressure, heart rate, and gait data were obtained at 0 (first sample), 2.5, 5, 7.5, and 10 hours. In the second phase, the same data were gathered immediately before (0 hours) and 2.5, 5, 7.5, and 10 hours after induction of acute urate synovitis in the left stifle joint. Data were statistically evaluated to compare changes in variable values over time and to determine the accuracy of serum cortisol measurements for diagnosis of acute orthopedic pain. Results-Following induction of synovitis, ground reaction forces were significantly decreased relative to preinduction values at 2.5, 5.0, 7.5, and 10.0 hours and serum cortisol concentration was significantly increased at 2.5 hours. A cortisol concentration of >/= 1.6 mug/dL indicated pain with a 91% sensitivity and 35% specificity. Conclusions and Clinical Relevance-In this model, cortisol concentration may be useful for diagnosing pain in dogs. Although, with a cutoff of >/= 1.6 mug/dL, pain would be detected in most dogs with pain, some pain-free dogs would also be identified as having pain. Conversely, dogs with a serum cortisol of < 1.6 mug/dL would be unlikely to have pain. Validation of this diagnostic test in a large, heterogeneous group of clinical patients is necessary.