To examine the validity of extrapolating parenteral product bioequivalence determinations across target animal species, the relative bioavailability of two injectable formulations of ampicillin trihydrate (PolyflexR, a water-based suspension, and Ampi-kel 10R, an oil-based suspension) was examined in calves, sheep and swine. Employing products recognized to be bioinequivalent provided an opportunity to explore potential species-by-formulation interactions. As compared with PolyflexR, Ampi-kel 10R exhibited lower area under the curve (AUC) estimates but higher peak concentrations in all target animal species. Nevertheless, marked interspecies differences were noted in the width and bounds of the confidence intervals about the differences in treatment means. Potential physiological and physico-chemical reasons for these findings are discussed.