Tracer amounts of dehydroepiandrosterone-4-C14 (DHA-4-C14) and dehydroepiandrosterone-7[alpha]-H3 sulfate (DHAS-7a -H3) were injected simultaneously into a uterine artery in 2 patients 12 to 15 minutes prior to therapeutic interruption of pregnancy at laparotomy and the amniotic fluid, fetuses, placentas and urine samples were analysed for radioactive material. No radioactivity was found in the amniotic fluid. Regardless of the tracer administered, approximately 80% of the radioactive material recovered from the placentas was in an unconjugated (free) form, whereas more than 80% of the radioactivity present in the fetuses was in a conjugated form. More than 67% of the total radioactive material in the 2 placentas and more than 87% of that in the 2 fetuses was present as phenolic material. Approximately half of this phenolic material was identified as estrone (OE1) and 17[beta]-estradiol (OE2). The H3 to C14 ratio of OE1 and OE2 isolated from the fetuses and urine specimens was lower than the ratio of the injected material. Corresponding H3 to C14 ratios in placental OE1 and OE2 were higher than those in fetal or urinary OE1 and OE2. Labelled estriol (OE3) could be detected in the placentas, but not in the fetuses. However. a 16-hydroxylated phenolic intermediate (probably 16[alpha]-hydroxyestrone) was found in 1 fetus studied. The results indicate that the DHA and DHAS of the maternal circulation can serve as direct precursors to placental OE2 and OE1 and as indirect precursors - via a fetal intermediate - to placental OE3. Both compounds reach the fetus mostly as phenolic material, the free form being a better precursor than the sulfate. It is suggested that - around midpregnancy at least -the human placenta functions as a barrier limiting the transfer to the fetus of androgens present in the maternal circulation.