Recent studies suggest that ischemic preconditioning (IP) of the lung may have a protective effect in ischemia-reperfusion (I/R) injury. The purpose of the present study was to investigate the preconditioning hypothesis in rat pulmonary vascular bed and to examine the role of nitric oxide (NO) in IP. Isolated rat lung was perfused with Krebs-Henseleit solution containing indomethacin at a constant flow rate and perfusion pressure changes was recorded by a pressure transducer. In rat pulmonary vascular bed, 2 hours of hypothermic ischemia significantly attenuated histamine-induced vasodilator responses without affecting sodium nitroprusside (SNP) vasodilation when compared to sham values. However, 2 cycles of 5 minutes of ischemia and reperfusion that were applied prior to 2 hours of ischemia (IP protocol) prevented the attenuation of histamine-induced vasodilation. On the other hand, IP failed to prevent pulmonary edema after ischemia. Histopathological examination of rat lungs demonstrated that IP was able to protect endothelial cells and type II pneumocytes in lung. Moreover, in IP group, malondialdehyde (MDA) contents of the lung tissue were significantly lower and tissue glutathione (GSH) contents were significantly higher than those in I/R group. Administration of NO synthase inhibitor, N-G-nitro-L-arginine-methyl ester (L-NAME) prior to the IP protocol abolished the protective effects of IP, but not affected the tissue malondialdehyde and glutathione levels. These results suggest that I/R impaired endothelium-dependent vasodilatory response, whereas endothelium-independent SNP-induced responses were preserved in rat pulmonary vascular bed. IP prevented the impairment of pulmonary vascular endothelium-dependent responses, and these effects may be partially mediated by NO.