The pharmacodynamics (PD) of oxytetracycline was investigated against a strain of Mannheimia haemolytica. In vitro measurements, comprising minimum inhibitory concentration (MIC), minimum bactericidal concentration and time-kill curves, were conducted in five matrices; Mueller Hinton Broth (MHB), cation-adjusted MHB (CAMHB) and calf serum, exudate and transudate. MICs were much higher in the biological fluids than in MHB and CAMHB. Ratios of MIC were, serum: CAMHB 19 : 1; exudate:CAMHB 16.1; transudate:CAMHB 14 : 1. Ex vivo data, generated in the tissue cage model of inflammation, demonstrated that oxytetracycline, administered to calves intramuscularly at a dose rate of 20 mg/kg, did not inhibit the growth of M haemolytica in serum, exudate and transudate, even at peak concentration. However, using in vitro susceptibility in CAMHB and in vivo-determined pharmacokinetic (PK) variables, average and minimum oxytetracycline concentrations relative to MIC (C(av)/MIC and C(min)/MIC) predicted achievement of efficacy for approximately 48 hours after dosing. Similar C(av)/MIC and C(min)/MIC data were obtained when oxytetracycline was administered in the presence of carprofen. PK-PD integration of data for oxytetracycline, based on MICs determined in the three biological fluids, suggests that it possesses, at most, limited direct killing activity against M haemolytica. These data raise questions concerning the mechanism(s) of action of oxytetracycline, when administered at clinically recommended dose rates.