We sought to reduce early ischemia-reperfusion injury after lung preservation by an initial brief period of hemodilute reperfusion. Left lungs of New Zealand White rabbits were ventilated with room air and reperfused in an ex vivo ventilation-perfusion apparatus after 18 hours of storage at 10 °C. Lungs were randomly assigned to one of three groups (n = 6) according to the composition of initial reperfusate. In group 1 (control), preserved lungs were reperfused with whole blood for 20 minutes (hematocrit, 38%). In the experimental groups, blood was diluted to a hematocrit of 10% with Ringer's lactate (group 2) or low-potassium-dextran solution (group 3) for the first 10 minutes of reperfusion, followed immediately by whole blood for 10 minutes. Oxygen tension of left ventricular effluent at the end of the 20-minute assessment period was significantly higher in both hemodiluted groups (mean ± standard error of the mean: group 2, 81.3 ± 6.6 mm Hg; group 3, 77.0 ± 9.5 mm Hg, versus Group 1, 46.3 ± 7.4 mm Hg; p < 0.006). Similarly, mean tracheal airway pressure was reduced in the hemodiluted groups, suggesting improved compliance (group 2; 3.1 ± 0.3 mm Hg; group 3, 2.8 ± 0.6 mm Hg; versus group 1, 6.5 ± 1.4 mm Hg; p < 0.05). An initial 10-minute period of hemodilute reperfusion appears to reduce early pulmonary ischemiareperfusion injury in this 18-hour ex vivo rabbit lung preservation model.