OBJECTIVE: The purpose of this study was to determine the degree of cardiac oxygenation produced by different routes of extracorporeal membrane oxygenation in fetal lambs submerged in warm saline solution. STUDY DESIGN: Seven fetal lambs ranging in age from 113 to 133 days of gestation were delivered by cesarean section and oxygenated with extracorporeal membrane oxygenation. To maintain the patency of the ductus arteriosus, prostaglandin E(1) was continuously infused intravenously to the fetus. Initially the extracorporeal membrane oxygenation route was from the right atrium to the carotid artery. Then the extracorporeal membrane oxygenation route was changed to flow from the right atrium to the umbilical vein. The fetus was kept in a warm saline solution bath, and the fetal circulation was maintained, Extracorporeal membrane oxygenation flow ranged between 100 and 200 ml/min throughout the experiment. Simultaneous blood samples were taken during both types of extracorporeal membrane oxygenation from the following points in the fetal circulation: premembrane (least oxygenated blood leaving the fetus from the right atrium), postmembrane (oxygenated blood returning to the fetus), the carotid artery, and the left ventricle. The respiratory gases and pH of each sample were measured. Six fetuses received nonradioactive colored microspheres injected into the oxygenated blood returning to the fetus flow before returning to the fetuses during both types of extracorporeal membrane oxygenation. After the animals were killed, microspheres were counted in the myocardium separately taken from the right and left atria and the right and left ventricles to determine cardiac blood flow. RESULTS: During right atrium to carotid artery extracorporeal membrane oxygenation, left ventricle Po-2 remained low as postmembrane Po-2 increased; these values were not significantly correlated (r = 0.234, p = 0.61). During right atrium to umbilical vein extracorporeal membrane oxygenation, left ventricle and postmembrane Po-2 exhibited a significant positive correlation (r = 0.855, p = 0.014). When the extracorporeal membrane oxygenation route was switched from the right atrium to carotid artery to the right atrium to umbilical vein, there was a significant increase in left ventricle Po-2 and a decrease in left ventricle Pco(2), whereas the respiratory gases and pH remained unchanged at other sites in the circulation. Microsphere counts were consistently higher during right atrium to umbilical vein extracorporeal membrane oxygenation than during right atrium to carotid artery extracorporeal membrane oxygenation in all four samples from different parts of myocardium (p < 0.001 by paired t test). CONCLUSION: More effective cardiac oxygenation is provided by right atrium to umbilical vein extracorporeal membrane oxygenation than by right atrium to carotid artery extracorporeal membrane oxygenation.