Six dogs were treated with a single intravenous (i.v.) dose (2 mg/kg) of marbofloxacin, followed by single oral (p.o.) doses of marbofloxacin at 1, 2 and 4 mg/kg, according to a three-way crossover design. The same experimental design was used for the subcutaneous (s.c.) route. In addition, a long-term trial involving eight dogs given oral doses of marbofloxacin at 2, 4 and 6 mg/kg/day for thirteen weeks was carried out. Plasma and urine samples were collected during the first two trials, plasma and skin samples were collected after the second of these trials. Plasma, urine and skin concentrations of marbofloxacin were determined by a reverse phase liquid chromatographic method. Mean pharmacokinetic parameters after i.v. administration were the following: t1/2β=12.4h; ClB= 0.10 L/h.kg; Varea= 1.9 L/kg. The oral bioavailability of marbofloxacin was close to 100% for the three doses. At 2 mg/kg, Cmax of 1.4 μg/mL was reached at tmax of 2.5 h. Mean AUC and Cmax values had a statistically significant linear relationship with the doses administered. About 40% of the administered dose was excreted in urine as unchanged parent drug. After s.c. administration, the calculated parameters were close to those obtained after oral administration, except tmax (about 1 h) which was shorter. The mean skin to plasma concentration ratio after the long-term trial was 1.6, suggesting good tissue penetration of marbofloxacin.