Summary:Four immunochemical methods for digoxin assay were used to analyse control samples, 33 amniotic fluid samples, 57 samples from digitalis-treated, non-pregnant women, 90 pregnancy serum samples, and 72 samples of fetal or neonatal serum with or without digoxin therapy. One hundred and five samples were also submitted to ultrafiltration before analysis.Three methods (RIA, TD X , AMERLITE) showed practically the same precision, while the precision of the DELFIA was markedly inferior. In the analysis of serum samples from digoxin-treated, non-pregnant women, RIA and TD X gave practically the same values, whereas AMERLITE and DELFIA gave significantly higher values.Pregnancy serum and fetal serum contain "digoxin-like immunoreactive factors", and the qualitative and quantitative effects of these interfering factors are different for each of the four methods. The greatest sensitivity to "digoxin-like immunoreactive factors" is shown by TD X and DELFIA, while the lowest interference by "digoxin-like immunoreactive factors" is found in the analysis of ultrafiltered samples, using the TD X method.The composition of the "digoxin-like immunoreactive factors" in pregnancy serum and in fetal serum is altered by digoxin therapy, and these changes have different effects on the various analytical methods. The concentration of "digoxin-like immunoreactive factors" in the serum of fetuses receiving digoxin is markedly lower than that of healthy fetuses.For the reliable monitoring of digoxin therapy in the maternal and fetal circulation, the blood samples must be submitted to ültrafiltration before analysis.