Concentrations of immunoreactive (ir-) inhibin in circulation, amniotic fluid, and testes of embryos and newly hatched ducks were determined from d 21 of incubation to d 1 of age. Plasma concentrations of FSH and LH were also determined by chicken radioimmunoassay (RIA) systems. In addition, gene expression and cellular source of inhibin were investigated by in situ hybridization and immunohistochemistry. The results showed that plasma ir-inhibin gradually declined from d 21 to d 24, followed by an increase on d 25 and remained high until d 1 after hatching. FSH in plasma was high on d 21 followed by a sharp decline toward d 25 after which FSH levels stabilized. A reverse relationship was observed between inhibin and FSH during the late stage of incubation. Embryonic testes contained high ir-inhibin levels. Testicular ir-inhibin levels were relatively high at early time points with a peak on d 23, and significantly decreased from d 23 to d 24 and stabilized thereafter. Amniotic fluid concentrations of ir-inhibin were relatively low and remained constant between d 21 and d 25. In situ hybridization demonstrated that the expression of inhibin alpha- and betaA-subunit mRNA was coexisted in the cells in the seminiferous tubules of testes on d 25. The immunoreactivity of inhibin betaA- and betaB-subunits was colocalized in the cells in the seminiferous tubules of testes on d 25. The results of dimeric inhibins determined by the ELISA method showed that inhibin B can be measured in embryonic testicular homogenate and pooled embryonic plasma. Although inhibin A was detected in testicular homogenate, it was under the detection limit in pooled embryonic plasma. In conclusion, these results indicate that cells in the seminiferous tubules of embryonic testes in ducks may secrete dimeric (bioactive) inhibins to circulation and that the FSH-inhibin feedback loop may become operational during the late stage of the incubation.