We have reported that hypoxia stimulates EDRF(s) release from endothelial cells and the release may be augmented by previous hypoxia. As a mechanism, it was hypothesized that reoxygenation can stimulate EDRF(s) release from endothelial cells and we tested the hypothesis via bioassay experiment. In the bioassay experiment, rabbit aorta with endothelium was used as EDRF donor vessel and rabbit carotid artery without endothelium as a bioassay test ring. The test ring was contracted by prostaglandin F2Ձ (3⁓10⁣6 M) which was added to the solution perfusing through the aorta. Hypoxia was evoked by switching the solution aerated with 95% O2/5% CO2 mixed gas to one aerated with 95% N2/5% CO2 mixed gas. Hypoxia/reoxygenation were interexchanged at intervals of 2 minutes (intermittent hypoxia). In some experiments, endothelial cells were exposed to 10-minute hypoxia (continuous hypoxia) and then exposed to reoxygenation and intermittent hypoxia. In other experiments, the duration of reoxygenation was extended from 2 minutes to 5 minutes. When the donor aorta was exposed to intermittent hypoxia, hypoxia stimulated EDRF(s) release from endothelial cells and the hypoxia-induced EDRF(s) release was augmented by previous hypoxia/reoxygenation. When the donor aorta was exposed to continuous hypoxia, there was no increase of hypoxia-induced EDRF(s) release during hypoxia. But, after the donor aorta was exposed to reoxygenation, hypoxia-induced EDRF(s) release was markedly increased. When the donor aorta was pretreated with nitro-L-arginine (10⁣5 M for 30 minutes), the initial hypoxia-induced EDRF(s) release was almost completely abolished, but the mechanism for EDRF(s) release by the reoxygenation and subsequent hypoxia still remained to be clarified. TEA also blocked incompletely hypoxia-induced and hypoxia/reoxygenation-induced EDRF(s) release. EDRF(s) release by repetitive hypoxia and reoxygenation was completely blocked by the combined treatment with nitro-L-arginine and TEA. Cytochrome P450 blocker, SKF-525A, inhibited the EDRF(s) release reversibly and endothelin antgonists, BQ 123 and BQ 788, had no effect on the release of endothelium-derived vasoactive factors. Superoxide dismutase (SOD) and catalase inhibited the EDRF(s) release from endothelial cells. From these data, it could be concluded that reoxygenation stimulates EDRF(s) release and hypoxia/reoxygenation can release not only NO but also another EDRF from endothelial cells by the production of oxygen free radicals.