It is now becoming increasingly clear that free radicals contribute to brain damage in several conditions, such as hyperoxia and trauma. It has been more difficult to prove that free radical production mediates ischemic brain damage, but it has often been suggested that it may be a major contributor to reperfusion damage, observed following transient ischemia. Recent results demonstrate that cerebral ischemia of long duration, particularly when followed by reperfusion, leads to enhanced production of partially reduced oxygen species, notably hydrogen peroxide (H2O2). It has also been suggested that postischemic hyperoxia, e.g. an increased oxygen tension during the recirculation period, adversely affects recovery following transient ischemia. Other data support the notion that brain damage caused by permanent ischemia (stroke) is significantly influenced by production of free radicals. The present study, however, fails to show that recirculation following brief periods of ischemia (15 min) leads to an enhanced H2O2 production, and that hyperoxia aggravates the ischemic damage. This study was undertaken to reveal whether variations in oxygen supply in the postischemic period following forebrain ischemia in rats affect free radical production and the brain damage incurred. To that end, rats ventilated on N2O/O2 (70:30) were subjected to 15 min of transient ischemia. Normoxic animals were ventilated with the N2O/O2 mixture, hyperoxic animals with 100% O2, and hypoxic ones with about 10% O2 (balance either N2O/N2 or N2) during the recirculation. At the end of this period, the animals were decapitated for assessment of H2O2 production with the aminotriazole/catalase method. This method is based on the notion that aminotriazole interacts with H2O2 to inactivate catalase; thus, the rate of inactivation of catalase in aminotriazole treated animals reflects H2O2 production. In a parallel series, animals ventilated with one of the three gas mixtures in the early recirculation period, respectively, were allowed to recover for 7 days, with subsequent perfusion-fixation of brain tissues and light microscopical evaluation of the brain damage. Animals given aminotriazole, whether rendered ischemic or not, showed a reduced tissue catalase activity, reflecting H2O2 production in the brain. Hyperoxic animals failed to show increased tissue H2O2 production, while hypoxic ones showed a tendency towards decreased production. However, all three groups (hypo, normo- and hyperoxic) had similar density and distribution of neuronal damage. These results suggest that although postischemic oxygen tensions may determine the rates of H2O2 production, variations in oxygen tensions do not influence the final brain damage incurred.(ABSTRACT TRUNCATED AT 400 WORDS)