SUMMARY A cross-over experiment was used to compare serum ampicillin concentrations in mares. Six mares and 2 dose regimens were used. Four of 6 mares were in midterm of gestation; the remaining 2 were not pregnant. Ampicillin was injected im and was administered every 12 hours for 3 doses. Dose regimens compared were sodium ampicillin (22 mg/kg) given alone, or sodium ampicillin (11 mg/kg) and ampicillin trihydrate (11 mg/kg) given concurrently. Blood samples were collected over 12 hours after the 1st and 3rd injections. Serum ampicillin concentrations were determined by bioassay. Peak mean serum ampicillin concentration occurred 0.5 hour after injection. Serum ampicillin concentrations after sodium ampicillin injection (22 mg/kg) were significantly higher than concentrations after the concurrent-dose regimen, for up to 4 hours (P < 0.05). Concentrations between 4 and 12 hours were not significantly different. Concentrations in both groups between 8 and 12 hours after injections were done were < 2 μg/mL The halflives of ampicillin in serum was 2.06 hours with the dosing with sodium ampicillin and 3.33 hours with the concurrent dosing of sodium ampicillin and ampicillin trihydrate. Peak serum ampicillin concentrations following sodium ampicillin injection were between 5.53 and 26.59 μg/ml. Great variation was noted between animals; however, little variation was noted within the individual. Peak serum concentrations following repeated doses within a given mare were not significantly different (P > 0.05). The mean peak serum ampicillin concentration of the pregnant mares (8.75 μg/ml) was lower than that of nonpregnant mares (21.57 μg/ml).