As estetrol (E4) is believed to be the steroid most likely to wholly dependent on fetal origin, we developed a radiommunoassay for unconjugated and conjugated E4 (E4-U and E4-G) and investigated plasma and urinary levels serially throughout the second half of pregnancy to establish their validity by means of monitoring or screening tests to assess fetal well-being. E4 exhibits a remarkable increase during the latter half of pregnancy. At term, the mean E4-U level in maternal peripheral plasma was 0.67 +/- 0.33 ng/ml, a five fold increase from that at 28 weeks; E4-G was 4.57 +/- 2.84 ng/ml, showing a four fold increase; and E4-G levels in maternal urine were 1.68 +/- 0.96 mg/day, showing a three fold increase from that at 28 weeks. E4-U and E4-G levels showed no diurnal change. The coefficient of the correlation between plasma E4-U and E4-G was 0.699, which is satisfactory, but no correlation was found between urinary and plasma E4 levels. A significant correlation was shown between maternal and umbilical E4-U (r=0.820) and E4-G (r=0.608). No relationships between E4 levels and birth weight were detected. Pre-eclampsia, Rh-isoimmunization and diabetes mellitus are common complications of pregnancy which may cause latent fetal distress. Prenatal fetal assessment was performed by serial daily evaluations of these E4 values. In pre-eclampsia resulting in a small full term baby, E4 levels were mostly below normal mean values or failed to show an increased pattern. In addition, the E4 levels decreased in one case of neonatal death. In Rh-isoimmunization, plasma E4-G levels were lower in the group affected severely by the hemolytic desease. In a patient with diabetes mellitus delivered of a healthy baby, E4 levels were within the range of a normal pregnancy. In order to evaluate fetal and placental reserve capacities as well as feto-placental function, the dehydroepiandrosterone sulfate (DHA-S) loading test was performed by loading selected subjects with 50 mg of DHA-S, then serially measuring the E4 in the maternal plasma and urine. Intravenous infusion of 50 mg of DHA-S was completed in 60 minutes. A rapid and sharp increase of plasma E4 was observed, reaching maximal concentrations at 120 minutes in normal pregnancies. However, urinary levels showed patterns similar to those reported for estriol (E3). In some abnormal pregnancies, no increased or delayed patterns were observed in plasma E4-G levels, while the serial levels remained within the normal range. This possibly suggests that in these pregnancies, fetal functions had been inhibited or had reached their limit. It is concluded that the simultaneous determinations of serial E4 levels accompanied by the DHA-S loading test may be of value in assessing fetal well-being and reserve capacity and may therefore improve fetal and neonatal prognosis in abnormal pregnancies.