The aim of this study was to evaluate the influence of different maternal and fetal albumin concentrations on the transplacental transfer and the placental tissue accumulation of digoxin. Digoxin passage across the isolated lobules of 15 human placentae was calculated from repeated fetal and maternal perfusate samples, and placental tissue digoxin concentrations were measured at the end of the experiments. Metildigoxin (Lanitop) was added to the maternal medium at a concentration of 5.70 +/- 0.73 ng mL-1, and maternal and fetal perfusate albumin (BSA) concentrations were kept equal either at a high concentration of 21 g L-1 (Group I; n = 5) or at a low concentration of 3 g L-1 (Group III; n = 5), or differed with a materno-fetal gradient of 21:3 g L-1 (Group II; n = 5). In the experiments with low maternal albumin concentrations (Group III), digoxin concentrations in the maternal circuit decreased to 3.56 ng mL-1, whereas digoxin concentrations in the fetal circuit reached 2.59 ng mL-1 over a 3-h period. With maternal BSA concentrations of 21 g L-1 (Group I and Group II), the decrease in digoxin concentration in the maternal circuit was lower (P < 0.05), and digoxin tissue concentrations at the end of the experiments were smaller (0.45 +/- 0.07 and 0.42 +/- 0.03 v. 0.82 +/- 0.32 ng mg-1 protein, Group I and Group II v. Group III respectively; P < 0.05). Comparing only those lobules with similar high concentrations of maternal protein, fetal BSA concentrations of 21 g L-1 resulted in a greater increase in digoxin concentrations in the fetal circuit (end-feto to initial-maternal digoxin concentrations of 0.44 +/- 0.08 v. 0.37 +/- 0.04 ng mg-1 protein (Group I v. Group II respectively), although this was not significant. The data suggest that maternal and fetal serum albumin concentrations may have an influence on transplacental digoxin transfer, and this should be considered when treating fetuses with cardiac disease transplacentally with glycosides.