Background: Subjective pain assessment scales have been widely used for assessing lameness in response to pain, but the accuracy of these scales has been questioned. To assess scale accuracy, 10 lame, presa Canario dogs with osteoarthritis (OA) associated with bilateral hip dysplasia were first treated with mesenchymal stem cells. Then, potential lameness improvement was analyzed using two pain scales (Bioarth and visual analog scale). These data were compared with similar data collected using a force platform with the same animals during a period of 6 months after treatment. Results: The F test for intraclass correlation showed that concordance in pain/lameness scores between the 2 measuring methodologies was not significant (P value ≥ 0.9213; 95 % confidence interval, –0.56, 0.11). Although subjective pain assessment showed improvement after 6 months, force platform data demonstrated those same animals had returned to the initial lameness state. Conclusion: Use of pain assessment scales to measure lameness associated with OA did not have great accuracy and concordance when compared with quantitative force platform gait analysis.