Hypoxia-reoxygenation (H/R) causes tissue injury, mainly due to free radical production and leukocyte activation. H/R-induced endothelial damage is widely described, however in pharmacological research, there are only sporadic functional studies investigating in vitro vascular H/R. This methodological study compares results of in vivo and in vitro functional experiments. In canine and porcine in vivo experiments hearts were subjected to regional or global ischemia and reperfusion. Blood flow was measured on the left anterior descending coronary artery with a perivascular ultrasonic probe. Endothelium-dependent and -independent vasodilation was assessed after single-bolus intracoronary administration of acetylcholine and sodium nitroprusside (SNP). In organ bath experiments, isolated porcine coronary and rat aortic rings were investigated. Hypoxia (30, 45, 60, 120 min) was induced in the chamber by gassing with 95% N2-5% CO2. (pO2 < 30 mm Hg) During the subsequent reoxygenation (30 min), gassing was changed to 95% O2-5% CO2. The dose-dependent vasoresponse to acetylcholine, bradykinin and SNP was investigated in precontracted rings under normoxic conditions and after H/R. Endothelial function assessed by coronary blood flow measurements was impaired after ischemia-reperfusion in vivo. Although the typical hypoxic vasomotor response could be observed in vitro, no impairment of endothelial function could be proven after H/R in any groups. We conclude that endothelial injury occurring in vessel rings during in vitro H/R is too slight (probably due to lack of activated leukocytes) and cannot be demonstrated in functional measurements. Therefore the experimental model of in vitro vascular H/R is not suited for reliable investigation of pharmacological attempts.