The objectives of this study were to examine the pharmacokinetics of tobramycin in the horse following intravenous (IV), intramuscular (IM), and intra-articular (IA) administration. Six mares received 4 mg/kg tobramycin IV, IM, and IV with concurrent IA administration (IV+IA) in a randomized 3-way crossover design. A washout period of at least 7 days was allotted between experiments. After IV administration, the volume of distribution, clearance, and half-life were 0.18 ± 0.04 L/kg, 1.18 ± 0.32 mL·kg/min, and 4.61 ± 1.10 h, respectively. Concurrent IA administration could not be demonstrated to influence IV pharmacokinetics. The mean maximum plasma concentration (Cmax) after IM administration was 18.24 ± 9.23 μg/mL at 1.0 h (range 1.0–2.0 h), with a mean bioavailability of 81.22 ± 44.05%. Intramuscular administration was well tolerated, despite the high volume of drug administered (50 mL per 500 kg horse). Trough concentrations at 24 h were below 2 μg/mL in all horses after all routes of administration. Specifically, trough concentrations at 24 h were 0.04 ± 0.01 μg/mL for the IV route, 0.04 ± 0.02 μg/mL for the IV/IA route, and 0.02 ± 0.02 for the IM route. An additional six mares received IA administration of 240 mg tobramycin. Synovial fluid concentrations were 3056.47 ± 1310.89 μg/mL at 30 min after administration, and they persisted for up to 48 h with concentrations of 14.80 ± 7.47 μg/mL. Tobramycin IA resulted in a mild chemical synovitis as evidenced by an increase in synovial fluid cell count and total protein, but appeared to be safe for administration. Monte Carlo simulations suggest that tobramycin would be effective against bacteria with a minimum inhibitory concentration (MIC) of 2 μg/mL for IV administration and 1 μg/mL for IM administration based on Cmax:MIC of 10.