Background: Mesenchymal stromal cells (MSCs) are widely used for osteoarthritis (OA) treatment. Optimizing dose, timing, and safety while comparing efficacy with standard therapies like hyaluronic acid (HA) is essential for their standardization. Objectives: To assess the safety and efficacy of eCB-MSCs in client-owned horses with fetlock or carpus OA. Study design: Prospective blinded randomized clinical study. Methods: Horses of any breed, age, sex, or sport horse discipline diagnosed by blocking with fetlock or carpus OA were randomly assigned to receive either 10 or 20 million eCB-MSCs in hyaluronic acid (HA), while the control group received 3 ml of HA alone. Subjective lameness examinations were performed at baseline, 3- and 6- weeks post-treatment; additionally, 24-72 hours post-injection, horses were evaluated for adverse reactions. Follow-up surveys were sent to owners 18-weeks after treatment. Results: A total of 27 client-owned horses were enrolled. No significant adverse reactions occurred. There were no significant differences in lameness outcomes between treatment groups (HA: 1.7 ± 1.24 [1.22-2.27]; 10-MSC+HA: 1.2 ± 1.27 [0.67-1.73]; 20-MSC+HA: 1.1 ± 1.28 [0.55-1.64] grades) (P > 0.05) or by treatment by time (P > 0.05). All groups, irrespective of treatment, improved by week 6 by 0.5 ± 0.65 [0.2-0.8] grades (P < 0.05). Although return-to-work rates were not significantly different between treatment groups (P > 0.05), both MSC groups had higher rates of return to the same or higher work levels compared to HA-only (89%, 78%, and 56%, respectively). Main limitations: This study aimed to assess the safety and efficacy of MSC treatment in client-owned horses. It expected higher exercise return rates at 18 weeks for MSC groups (10 or 20 million). However, unexpectedly high returns in the HA group rendered the study underpowered. Based on observed rates, 30 horses per group would be needed for confident conclusions favouring MSCs.