After the administration of racemic ketoprofen and carprofen to man, both enantiomers of each compound exhibit similar plasma profiles. This contrasts with the rat where the active S(+) enantiomer is predominant. For carprofen, regardless of the route of administration, the R(-) enantiomer is predominant in the plasma of all investigated animal species. The S(+)/R(-) ratio of the "areas under the curves" during the time course of the kinetics, is: 0.60 in dogs, 0.53 in Yucatan micro-pigs, 0.48 in mini-goats, 0.67 in calves and 0.19 in horses. For ketoprofen, the S(+) enantiomer is predominant in dogs, cats and horses, with ratios of 30.3, 5.3 and 1.5, respectively, while R(-) is the predominant enantiomer in sheep. The interpretation of these inter-species differences can be supported by experimental evidence, however some informations are lacking and additional investigation is required. In the case of ketoprofen where S(+) is predominant in rats, dogs and horses, the metabolic chiral inversion from R(-) to S(+), which has been demonstrated in rats, may also take place in the latter two species. In addition, the well documented stereoselective clearance of the glucuronides, possibly in favour of the enantiomer S(+), may explain the lower body clearance of the R(-) enantiomer in sheep. For carprofen, no metabolic chiral inversion was shown in rats and dogs after administration of each enantiomer individually, but for this compound, stereoselective clearance of glucuronides has been demonstrated which may support the idea of a plasma concentration shift of the enantiomeric proportions vs time in favour of the R(-) enantiomer. Regardless of the possible biological mechanisms which are responsible for these inter-species differences, the existence of these differences gives rise to at least two important issues: The choice of animal species which can be used in the research of drugs destined for human therapeutics: the most pertinent animal species will be the one which demonstrates an enantiomeric plasma profile closest to that observed in man. The present data show that the ideal animal species from this respect has still to be identified. For application in veterinary therapeutics, a careful balance must be established between the requirement of favourable bioavailability of the active S(+) enantiomer and the potential of any possible chiral inversion of R(-) to generate hybrid molecules in meat and milk which in turn may lead to residues, the toxicity of which to the human consumer is still unknown.