Much greater quantities of 16α-hydroxyoestrogens (e.g. oestriol) than of 16-deoxyoestrogens (e.g. oestradiol-17β) are formed in human pregnancy than might be expected from the relative availability to the placenta of the 16α-hydroxy- and 16-deoxy-C 19 precursors. To investigate this further, 16α-hydroxyandrostenedione (16α-OH-A 4) and androstenedione (A 4) were tested in vitro as substrates and mutual inhibitors of human placental aromatase. It was found that the K m for aromatisation of A 4 (mean = 0.26 μmol/l) was very similar to K i (0.30, 0.35 μmol/l) for the inhibition by A 4 of the aromatisation of 16α-OH-A 4. Similarly, K m for aromatisation of 16α-OH-A 4 (mean = 1.21 μmol/l) had the same value as the K i (1.10, 1.2 μmol/l)for the inhibition by 16α-OH-A 4 of the aromatisation of A 4. From graphical analysis of Lineweaver-Burk plots, both inhibitions were characterised as noncompetitive. Hence, it was concluded that the two 16-deoxy- and 16-hydroxy-C 19, substrates bind at separate, but interactive, sites and that each substrate on binding inhibits the aromatisation of the other. Additional evidence for the separate but interactive substrate binding sites for the 16-deoxy- and 16-hydroxy-C 19 steroids was obtained by use of the suicide inhibitor 4-hydroxyandrostenedione (4-OH-A 4), which is recognised as binding to the aromatisation site for A 4. Aromatisation of 16α-OH-A 4 was found to be inhibited by pre-incubation of the microsomes with 4-OH-A 4 (0.1 μmol/l). The presence of A 4 (4.6 μmol/l), but not of 16α-OH-A 4 (4.0 μmol/l) during the pre-incubation successfully protected the subsequent aromatisation of 16α-OH-A 4 from this inhibition. In addition, the K m values, reported here, suggest also that the 16-deoxyandrogens are preferred to the 16α-hydroxyandrogens as oestrogen precursors. In consequence, factors other than substrate affinity and plasma concentrations must be presumed to be involved in the overwhelming production of 16α-hydroxyoestrogens in human pregnancy.