We have investigated the serum concentration of 16-dehydropregnenolone (3 beta-hydroxy-5,16-pregnadien-20-one) 3-sulfate (16-DHPS) in the umbilical artery (U.A.), umbilical vein (U.V.) and maternal vein (M.V.) to discover the origin of 16-DHPS. Although there was no significant difference between the levels of 16-DHPS in U.A. (18 +/- 15 ng/ml, mean +/- SD., n = 28) and U.V. (10 +/- 9 ng/ml, n = 28), these values were significantly higher (U.A., p < 0.001; U.V., p < 0.001) than that in M.V. (2 +/- 3 ng/ml, n = 28). These levels in the U.A. and U.V. did not fall in infants (30 +/- 18 ng/ml, n = 7) during the early neonatal period (2-7 d after birth). A significant correlation between the serum levels of 16-DHPS and 16-hydroxypregnenolone (3 beta, 16 alpha-dihydroxy-5-pregnen-20-one) 3-sulfate (16-OH-PregS), which may be the precursor steroid for 16-DHPS, was observed in the U.A. (r = 0.630, n = 28, p < 0.001), but not in the U.V. Moreover, this significant correlation persisted during the early neonatal period (p < 0.05, r = 0.842, n = 7), although the neonate had been separated from the maternal milieu. These results suggest that 16-DHPS originates in the fetus. To confirm the metabolic pathway of 16-DHPS (i.e. pregnenolone (3 beta-hydroxy-5-pregnen-20-one) 3-sulfate (PregS)-->16-OH-PregS-->16-DHPS), we investigated the correlation between the serum concentrations of the precursor steroid and the product in both the U.A. and U.V. A significant correlation was obtained between the serum concentrations of PregS and 16-OH-PregS both in the U.A. (p < 0.001, r = 0.563, n = 28) and U.V. (p < 0.05, r = 0.476, n = 27). As described above, the serum levels of 16-DHPS and 16-OH-PregS only correlated significantly in the U.A. These findings support the existence of the pathway, PregS-->16-OH-PregS-->16-DHPS, in the fetus.