For decades, superoxic ex vivo dual perfusion of the human placental lobule has been used as a model to study the physiology and metabolism of the placenta. The aim of this study was to further develop the technique to enable perfusion at soluble oxygen concentrations similar to those in normal pregnancy (normoxia) and in pre-eclampsia (PE; hypoxia). Our design involved reducing the mean soluble oxygen tension in the maternal-side intervillous space (IVS) perfusate to 5-7% and o3% for normoxia and hypoxia, respectively, while providing a more ubiquitous delivery of perfusate into the IVS, using 22 maternal-side cannulae. We achieved quasi-steady states in [O 2 ] fetal venous (soluble) , which were statistically different between the two adaptations at t ¼ 150 to t ¼ 240 min of dual perfusion (2.1, 1.2, 2.8 and 0.4, 0.0, 1.5%; median, 25th, 75th percentiles, n ¼ 20 and 24 readings in n ¼ 5 and n ¼ 6 lobules, normoxic and hypoxic perfusion, respectively; Po0.001, Mann-Whitney U-test). Lactate dehydrogenase (LDH) levels in fetal and maternal venous outflow perfusates were unaffected by the adaptations. There was also no difference in tissue lactate release between the two adaptations. Glucose consumption from the fetal circulation and maternal-side 'venous' pyruvate release were higher under normoxic conditions, indicative of a greater metabolic flux through glycolysis. Furthermore, there was greater release of the hypoxic-sensitive marker, macrophage inflammatory protein-1a, into the maternal venous perfusate in the hypoxic model. Also, during hypoxic perfusion, we found that fetal-side venous placental growth factor (PlGF) levels were higher compared with normoxic perfusion. We conclude that these ex vivo adapted methods of placental perfusion provide a means of studying aspects of placental metabolism in relation to normal oxygenation and hypoxia-associated pregnancy disease.