In order to study the role of neutrophils in hyperoxic lung edema, we induced a sustained granulocytopenia in rats by giving them a single intraperitoneal injection of 50 mg · kg”1 cyclophosphamide (CY) (CY-injected rats). Hyperoxic lung damage (FIO2 = 1.0, 60 h) was assessed by measuring pulmonary water (PW). Exposure to hyperoxia coincided with the granulocytopenic period. After exposure to hyperoxia, the difference in PW between CY- and 0.09% saline solution-injected (S-injected) rats was not significant (PW (g) (x ± 1 SD): CY-injected rats: 1.49 ± 0.09; S-injected rats: 1.68 ± 0.09, NS). If neutrophils have a pathogenic role in hyperoxic lung injury, the absence of edema reduction in neutropenic rats could be explained by pulmonary infection, insufficient neutrophil depletion or cumulative lung toxicity of CY and oxygen. No pulmonary infectious agents could be detected in lung and bronchoalveolar lavage cultures. Although already severe, the neutrophil depletion was intensified by a second injection of CY (50 mg · kg”1). After exposure to hyperoxia, the number of neutrophils in lung lavage was much lower in these neutropenic rats (156 · 103 ± 33 · 103) than in S-injected rats (571 103 ± 100–103) (p < 0.001), but no significant reduction of PW was observed. No significant increase was observed in either PW or dry lung weight in CY-injected rats not exposed to hyperoxia. CY alone might not be responsible for the absence of edema reduction, but might amplify the toxic effects of oxygen on the lung. In conclusion, we could not demonstrate any reduction of lung edema due to granulocyte depletion. The role of neutrophils in hyperoxic lung injury has yet to be fully elucidated.