Pharmacokinetic variables of skeletal muscle creatine kinase were determined in sheep after intravenous and intramuscular administration of the semipurified enzyme. Catheters implanted in the jugular vein were used for both intravenous injections and blood withdrawals. Blood sample collection by vacutainer and hemolysis may in fact have considerable effects on the measurement of creatine kinase activity in plasma. The change in the enzyme activity versus time in the plasma, after intravenous administration (123 +/- 38 U/kg) of creatine kinase, was fitted by a biexponential model. The mean volume of the central compartment (45 +/- 5 mL/kg) was approximately equal to the plasma volume. Plasma half-life and plasma clearance of creatine kinase were 3.7 +/- 1.7 h and 23 +/- 8 mL.kg-1.h-1, respectively. Mean plasma bioavailability of creatine kinase after intramuscular administration (357 +/- 36 U/kg) in both the loins and the gluteal mass was 42%. Maximal plasma activity was observed 4 and 5 h after injection and the half-life of the terminal phase was 7.3 or 8.6 h according to the muscle. The extent of muscle damage after intramuscular administrations of 21 veterinary drug formulations (one product per animal) was estimated from the total creatine kinase activity released in plasma during the 72 h following the injection. Equivalent weights of damaged muscle ranged from 1.4 to 83.3 g according to the irritant potency of the test formulation. Results differed only moderately between the injection sites (right and left gluteal mass) in the same animal. It can be concluded from this study that, in sheep: i) the bioavailability of creatine kinase from different injection sites (gluteal mass and loins) is comparable; and ii) the intra-individual variability in the estimation of muscle damage is moderate. Once validated, this non-invasive approach for local tolerance studies could be of value in assessing and comparing the irritant potency of veterinary drugs and in reducing the number of animals required.