The placental transfer of ranitidine was studied at pharmacokinetic steady state in anesthetized, full-term, pregnant sheep. Ranitidine was administered to the ewe in three preparations and to the fetus in three other sheep. In all experiments, dose size was based on the combined maternal-fetal weight. Steady-state plasma levels were reached within 4 hr by using an initial intravenous bolus dose followed by continuous infusion. Following maternal dosage, the mean maternal (jugular vein) steady-state concentration (CMss) at 4 hr was 842 +/- 66 ng/ml (SEM), the mean fetal (carotid artery) steady-state concentration (CFss) was 26.5 +/- 4.2 ng/ml, and the mean fetal umbilical venous steady-state concentration was 28.9 +/- 3.5 ng/ml. Both the fetal and umbilical plasma concentrations were significantly less than the maternal plasma concentrations (p less than 0.01). With fetal dosage, mean CMss was 414 +/- 42 ng/ml at 4 hr and was significantly less than the mean CFss value at the same time, which was 6890 +/- 360 ng/ml (p less than 0.005). Ranitidine was not bound extensively to plasma proteins in the ewe or the fetus (range 12-55% bound). The reversal of the CMss/CFss gradient with the change from maternal to fetal administration and the low binding of the drug shows that the gradient following maternal dosage cannot be explained by ion-trapping or differential plasma protein binding. As active placental transport is considered unlikely, the low fetal plasma concentrations are probably due to the presence of significant fetal elimination of ranitidine. Furthermore, the substantial gradient between maternal and umbilical venous plasma concentrations suggests that placental elimination of ranitidine should also be considered.