The clinical application of lung transplantation is severely limited by the shortage of suitable donor organs. Current techniques of lung preservation allow a maximum of 4 to 6 hours of safe ischemic time. The function of canine left lung allografts stored for 12 hours after being cooled by pulmonary artery flush was studied. Two types of flush solution were used: group I; Euro-Collins solution; group II, low-potassium-dextran solution. Lung function was studied immediately and 3 days after transplantation. This protocol enables study of acute preservation-related lung injury and the delayed manifestations of ischemic and reperfusion injury after a 3-day period of recovery. Inflatable cuffs were placed around each pulmonary artery at operation and were attached to subcutaneous injection ports. Temporarily occluding either pulmonary artery allowed independent study of the native or transplanted lung. Using this model, we were able to demonstrate reliable and reproducible preservation of lungs for 12 hours. The low-potassium-dextran solution provided significantly better immediate function of the preserved lung than the Euro-Collins solution: arterial oxygen tension 509 ± 15 mm Hg versus 111 ± 16 mm Hg (p < 0.0001). Function on the third day was excellent for both groups. Pulmonary artery pressure, pulmonary vascular resistance, and carbon dioxide tension were not significantly different between the groups immediately or on day 3.