The underlying biological mechanism of prenatal stress in humans is poorly understood, but maternal cortisol (F) excess seems to play an important role. In pregnant rats, acute stress causes an up-regulation of placental 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2), an enzyme present throughout the body (e.g., placenta, salivary glands) that inactivates F to cortisone (E), thereby protecting the fetus from maternal F overexposure. Whether acute stress influences human 11β-HSD2 is unclear. We aimed to explore the association between the maternal stress reactivity and amniotic fluid F, E, and the E/(E + F) ratio as a marker of fetoplacental 11β-HSD2. The predictive value of all markers for birth outcome was investigated. We examined 34 healthy pregnant women undergoing amniocentesis, which served as a standardized, real-life stressor. F, E, and E/(E + F) were determined from a single aliquot of amniotic fluid, and from saliva samples collected repeatedly. Subjects filled out state questionnaires repeatedly and were re-examined in a control condition after notification of a normal amniocentesis result. During amniocentesis, psychological stress, salivary F (SalF), and salivary E (SalE) increased significantly, whereas SalE/(E + F) decreased. The SalF reactivity was positively associated with amniotic E, while SalE/(E + F) was inversely associated with amniotic E/(E + F). SalF and SalE predicted lower and SalE/(E + F) higher birth weight. Psychological and amniotic fluid variables were unrelated to birth outcome. Findings indicate that maternal F is inactivated to E in the human fetoplacental unit during acute stress. Increased 11β-HSD2 activity within the maternal salivary glands following acute stress may mirror further stress protective mechanisms worthwhile investigating.