Amoxycillin was administered to six dogs intravenously (as the sodium salt at 20 mg/kg bodyweight) and orally (as the trihydrate at 20 mg/kg). The oral treatments followed a Latin square pattern, each dog receiving amoxycillin as a 60 ml suspension by stomach tube, or as 3 ml of drops or in the form of tablets. The concentration of the drug in the plasma was measured microbiologically and its pharmacokinetic parameters were calculated by the use of statistical moments. After intravenous administration the mean +/- sd apparent volume of distribution was 0.312 +/- 0.102 litre/kg, the steady state rate of clearance was 3.4 +/- 1.1 ml/min/kg and the mean residence time was 1.6 +/- 0.4 hours. After oral administration the liquid forms of the drug tended to be more readily absorbed than the tablets, as indicated by their higher bioavailabilities (suspension 76.8 +/- 16.7 per cent, drops 68.2 +/- 25.8 per cent, tablets 64.2 +/- 17.9 per cent). However, the differences between their pharmacokinetic parameters were not statistically significant. The respective values of Cmax for the tablets, drops and suspension were 18.6 +/- 5.3 micrograms/ml, 18.1 +/- 2.4 micrograms/ml and 20.7 +/- 2.2 micrograms/ml, of tmax 2.0 +/- 1.0 hours, 1.4 +/- 0.6 hours and 1.4 +/- 0.5 hours and of the AUC 69.5 +/- 22.5 micrograms/ml hours, 71.8 +/- 21.0 micrograms/ml hours and 80.6 +/- 21.8 micrograms/ml hours. The two useful drug products (drops and tablets) had similar pharmacokinetic profiles in the dogs and can therefore be regarded as equivalent in this species.