Paraquat poisoning and hyperbaric oxygen exposure are well established models of oxidative stress in lung. The aim of this study was focused on the contribution of oxygen free radicals and other cytotoxic species, such as lipid hydroperoxides, to the overall toxicity. Adult Wistar rats were injected with paraquat (30 or 60 mg/kg b.w.) or exposed to hyperbaric oxygen (0.2 MPa), and several parameters of lung damage were measured. Both treatments resulted in increased spontaneous lung chemiluminescence, number of lung PMN, malondialdehyde content, lung edema, and pleural liquid. Of note, spontaneous lung chemiluminescence, used to monitor the steady-state concentration of oxygen free radicals in vivo, did not increase significantly after either treatment. The increase in spontaneous lung chemiluminescence started after PMN migration, being both maxima separated by a delay time of 4–6 h. After PMN migrated and became activated in the lung, the survival of the animals started to decline. Thus, PMN can be considered as additional sources of oxygen free radicals supported by the subsequent increase in chemiluminescence. Their role in lung damage was evidenced by the increase in lung edema, augmented pleural liquid, and decreased survival after PMN migration. Lipid hydroperoxide concentration in lung membranes was also increased after either treatment. This increased concentration may be a consequence of an increased rate of lipid peroxidation, initiated by oxidative stress on lipid membranes, or by an inhibition of their catabolism. Ester lipid hydroperoxides normally produced in membranes cannot be catabolized directly by the glutathione peroxidase-reductase system unless phospholipase A 2 catalyses the release of free lipid hydroperoxides. In both experimental models, phospholipase A 2 activity was decreased to almost negligible values. Betamethasone (1 mg/ml, IV) administered to the rats 3 h before paraquat injection accelerated the decrease in survival and phospholipase A 2 inactivation. Inactivation of phospholipase A 2, detected in paraquat or oxygen exposed rats, could be attributed to a O 2 ·−-driven Fenton reaction. However, phospholipase A 2 inactivation by betamethasone pretreatment may be attributed to the presence of lipocortin, a corticosteroid-inducible factor and inhibitor of phospholipase A 2. Besides the mechanism underlying the inactivation of phospholipase A 2, the increase in lipid hydroperoxides may indicate their role as long-lived cytotoxic species that contribute to the damage already inititated by oxidative stress. Indeed, lipid hydroperoxides are very well known modifiers of membrane physical properties.