The technique of conducting high tidal volume (TV) ventilation-induced lung inflammation including remote organs is still open to discussion, and our aim is to investigate this issue in isolated ventilated rat lungs perfused with salt solution. Selective right lung (RL) hyperventilation (TV of 15 mL/kg with air containing 5% CO(2) on zero or 2.5 cm H(2)0 end expiratory pressure [ZEEP or PEEP] in addition to left lung (LL) on 2.5 cm H(2)0 continuous positive airway pressure (CPAP) for 60 min, was realized after 30 min both lungs ventilation by occluding the left main bronchus, and it was allocated to the following 5 groups: groups 1 and 2 underwent hyperventilation under ZEEP, groups 3 and 4 underwent hyper ventilation under PEEP with recirculation or nonrecirculation (R-ZEEP or NR-ZEEP and R-PEEP or NR-PEEP), and group 5 served as the control group. Recirculation means the same perfusate recirculates the system throughout the procedure. The wet/dry ratio and protein content of bronchoalveolar lavage fluid (Prot-BALF), cytokine messenger RNAs (mRNAs), localization of tumor necrosis factor-alpha (TNF-alpha) by immunofluorescence double staining, and TNF-alpha concentration in the perfusate and BALF in each lung were measured and compared between groups by Kruskal-Wallis test. Lung injury (increased wet/dry ratio, Prot-BALF, and TNF-alpha on endothelial and epithelial cells) was shown in the hyperventilated RLs with ZEEP compared with their corresponding CPAP LLs. PEEP prevented these injuries. Lung injury was also demonstrated in the recirculated LL compared with the nonrecirculated LL (Prot-BALF, TNF-alpha and interleukin-1beta [IL-1beta] mRNAs: the LL of the R-ZEEP is greater than the LL of NR-ZEEP by P < 0.01). Unilateral hyperventilated lungs with ZEEP induced TNF-alpha, increased permeability, and injured the control lung via perfusion.