The objectives of this study were to compare the pharmacokinetics and COX selectivity of three commercially available formulations of firocoxib in the horse. Six healthy adult horses were administered a single dose of 57 mg intravenous, oral paste or oral tablet firocoxib in a three-way, randomized, crossover design. Blood was collected at predetermined times for PGE2 and TXB2 concentrations, as well as plasma drug concentrations. Similar to other reports, firocoxib exhibited a long elimination half-life (31.07 ± 10.64 h), a large volume of distribution (1.81 ± 0.59L/kg), and a slow clearance (42.61 ± 11.28 mL/h/kg). Comparison of the oral formulations revealed a higher Cmax, shorter Tmax, and greater AUC for the paste compared to the tablet. Bioavailability was 112% and 88% for the paste and tablet, respectively. Maximum inhibition of PGE2 was 83.76% for the I.V. formulation, 52.95% for the oral paste formulation, and 46.22% for the oral tablet formulation. Pharmacodynamic modeling suggests an IC50 of approximately 27 ng/mL and an IC80 of 108 ng/ mL for COX2 inhibition. Inhibition of TXB2 production was not detected. This study indicates a lack of bioequivalence between the oral formulations of firocoxib when administered as a single dose to healthy horses.