Cefoxitin pharmacokinetics and bioavailability were studied in unweaned calves. The antibiotic was administered to nine calves intravenously (i.v.), to seven calves intramuscularly (i.m.) at 20 mg/kg and to eight calves i.m. at 20 mg/kg together with probenecid at 40 mg/kg. Serum concentration versus time data were analysed using statistical moment theory (SMT). The i.v. data were also fitted by a linear, open two-compartment model. The elimination half-life (t1/2) was 66.9 +/- 6.9 min (mean +/- SD) after i.v. and 81.0 +/- 10.9 min after i.m. administration. The t1/2 increased to 125.5 +/- 15.6 min by the co-administration of probenecid. The total body clearance (ClT) was 4.88 +/- 1.71 ml/min/kg and the volume of distribution (Vss) 0.3187 +/- 0.0950 l/kg. The mean residence time (MRT) was 68.2 +/- 12.3 min after i.v. and 118.6 +/- 16.8 min after i.m. injection and increased to 211.5 +/- 16.8 min by the co-administration of probenecid. The mean absorption time (MAT) was 50.6 min and the estimated bioavailability (F) of cefoxitin after i.m. administration was 73.8%. The cefoxitin protein binding ranged from 55.0 to 42.0% at concentrations from 2 to 50 micrograms/ml. The MIC90 values for cefoxitin were 6.25 micrograms/ml for E. coli and Salmonella group B isolates, 3.13 micrograms/ml for Salmonella group C and D and Pasteurella multocida. There were no statistically significant differences between the pharmacokinetic parameters calculated by SMT or compartmental analysis. SMT provided an additional independent parameter, the MRT, for characterization of drug disposition kinetics.