Comparative drug disposition studies can be useful in extrapolating from animals to man provided that the criteria indicating interspecies similarity in disposition reflect similar exposure to the foreign compound. Interspecies variability, on the other hand, can often be related to physiological or biochemical differences, thereby providing a rationale for the unsuitability or limitations of a species as a model for human metabolism. Retrospective evaluation of the following examples illustrates the relevance of the indicated disposition characteristics to risk and efficacy assessment: (a) oxaprozin (route of excretion, enterohepatic circulation and exposure; plasma concentrations and efficacy prediction); (b) ciramadol (species differences in presystemic elimination and major metabolic pathway); (c) acebutolol (pharmacologically active human metabolite absent in one of the toxicology species); (d) esmolol (duration of pharmacologic effect controlled by species dependent nature of blood esterases). Stereochemical preferences in the disposition of racemic drugs often differ among species. Extrapolations from one species to another cannot be made in this situation. Pharmacokinetic parameters based on measurements of the sum of the isomers are meaningless and potentially misleading. Future improvements can come from: computer assisted predictions of metabolic pathways; increased use of human tissues; and use of animal species physiologically similar to humans, e.g. the miniature swine.