To investigate whether respiratory acidosis modulates ventilator-induced lung injury (VILI), we perfused (constant flow) 21 isolated sets of normal rabbit lungs, ventilated them for 20 min (pressure controlled ventilation [PCV] = 15 cm H(2)O) (Baseline) with an inspired CO(2) fraction adjusted for the partial pressure of CO(2) in the perfusate (PCO(2) approximately equal to 40 mm Hg), and then randomized them into three groups. Group A (control: n = 7) was ventilated with PCV = 15 cm H(2)O for three consecutive 20-min periods (T1, T2, T3). In Group B (high PCV/normocapnia; n = 7), PCV was given at 20 (T1), 25 (T2), and 30 (T3) cm H(2)O. The targeted PCO(2) was 40 mm Hg in Groups A and B. Group C (high PCV/hypercapnia; n = 7) was ventilated in the same way as Group B, but the targeted PCO(2) was approximately equal to 70 to 100 mm Hg. The changes (from Baseline to T3) in weight gain (Delta WG: g) and in the ultrafiltration coefficient (Delta K(f) = gr/min/ cm H(2)O/100g) and the protein and hemoglobin concentrations in bronchoalveolar lavage fluid (BALF) were used to assess injury. Group B experienced a significantly greater Delta WG (14.85 +/- 5.49 [mean +/- SEM] g) and Delta K(f) (1.40 +/- 0.49 g/min/cm H(2)O/100 g) than did either Group A (Delta WG = 0.70 +/- 0.43; Delta K(f) = 0.01 +/- 0.03) or Group C (Delta WG = 5.27 +/- 2.03 g; Delta K(f) = 0.25 +/- 0.12 g/min/cm H(2)O/ 100 g). BALF protein and hemoglobin concentrations (g/L) were higher in Group B (11.98 +/- 3.78 g/L and 1.82 +/- 0.40 g/L, respectively) than in Group A (2.92 +/- 0.75 g/L and 0.38 +/- 0.15 g/L) or Group C (5.71 +/- 1.88 g/L and 1.19 +/- 0.32 g/L). We conclude that respiratory acidosis decreases the severity of VILI in this model.