Pharmacokinetic and pharmacodynamic parameters were established for enantiomers of the non-steroidal anti-inflammatory drug (NSAID) ketoprofen (KTP), each administered separately at a dose level of 1.1 mg/kg to a group of six New Forest geldings, in a three-period cross-over study using a tissue cage model of inflammation. For both S(+)-and R(-)-KTP, penetration into tissue cage fluid (transudate) and inflamed tissue cage fluid (exudate) was rapid, and clearances from exudate and transudate were much slower than from plasma. AUC values were, therefore, higher for exudate and, to a lesser degree, transudate than for plasma. Unidirectional chiral inversion of R(-)-to S(+)-KTP was demonstrated. Administration of both enantiomers produced marked, time-dependent inhibition of synthesis of serum thromboxane B2 and exudate prostaglandin E2, indicating non-selective inhibition of cyclo-oxygenase (COX) isoenzymes COX-1 and COX-2 respectively. Administration of both enantiomers also produced partial inhibition of beta-glucuronidase release into inflammatory exudate and of bradykinin-induced skin oedema. It is suggested that, although S(+)-KTP is generally regarded as the eutomer, R(-)-KTP was probably at least as active in inhibiting bradykinin swelling. Pharmacokinetic/pharmacodynamic (PK/PD) modelling of the data could not be undertaken following R(-)-KTP administration because of chiral inversion to S(+)-KTP, but pharmacodynamic parameters, Emax, EC50, N, keO and t1/2(keO). were determined for s(+)-KTP using the sigmoidal Emax equation. PK/DP modelling provided a novel means of comparing and quantifying several biological effects of KTP and of investigating its mechanisms of action.