We studied the effect of a vasodilator (prostaglandin E1) as well as flush (F) and storage (S) temperatures (4-degrees-C or 10-degrees-C) on lung preservation in an isolated rabbit lung perfusion model. Low-potassium dextran (LPD) or Euro-Collins (E-C) solution was used as flush solution. Six groups of six animals were studied: group 1 (LPD, 4-degrees-C F-S), group 2 (LPD with PGE1, 4-degrees-C F-S), group 3 (E-C with PGE1, 4-degrees-C F-S), group 4 (LPD, 10-degrees-C F-S), group 5 (LPD with PGE1, 10-degrees-C F-S), group 6 (E-C with PGE1, 10-degrees-C F-S). After 18-hr preservation, left lungs alone were ventilated, and reperfused with fresh venous blood. PaO2, PaCO2, pulmonary artery pressure (PAP), tracheal pressure (P(t)) during reperfusion, and wet/dry weight (W/D) ratios were measured. PaO2 after LPD with or without PGE1 was significantly higher than after E-C with PGE1 at 4-degrees-C (95.8 +/- 11.5 mmHg in group 1 or 102.7 +/- 8.6 in group 2 vs. 41.8 +/- 10.5 in group 3, P < 0.01) and at 10-degrees-C (119.3 +/- 2.3 in group 4 or 131.1 +/- 6.2 in group 5 vs. 54.6 +/- 5.2 in group 6, P < 0.01). PaCO2, PAP, P(t), and W/D ratios in the LPD groups were lower than in the E-C groups. LPD/PGE1 and LPD alone produced similar pulmonary preservation. PaO2 of lungs flushed with LPD and preserved at 10-degrees-C was higher than that of lungs stored at 4-degrees-C. We conclude that LPD solution is superior to E-C solution in this ex vivo rabbit lung preservation model, even when PGE1 is used. A moderate dose of PGE1 did not improve the performance of LPD as a flush solution. Pulmonary preservation with LPD at 10-degrees-C is superior to preservation at 4-degrees-C.