Endothelial cells (EC) cover the entire inner surface of the vessel system. They function as a selective barrier to protect the surrounding tissue form extravasations. Failure of this endothelial barrier caused by ischemia leads to oedema. During ischemia a rise in cytosolic calcium and a formation of gaps can be observed. These processes are further aggravated during the early period of reperfusion (‘reperfusion injury’). Former studies have shown that removal of extracellular calcium reduces the additional increase, but the calcium cannot be reduced to its original level. Could this be due to an internal release mechanism of calcium? To address this question, we monitored with optical imaging the cytosolic calcium level in living cardiac endothelial cells during reperfusion with a focus on the main intracellular store, the endoplasmatic reticulum. Microvascular coronary EC from humanely killed rats were exposed to simulated ischemia (40 min, pH 6.4) followed by reoxygenation (40 min, pH 7.4, 2.5 mM glucose). Cai was monitored via fura-2 fluorescence. The SERCA-inhibitor thapsigargin (THG, 10nM), the IP3-release-channel-inhibitor xestospongin C (XeC, 3μM), the phospholipase C inhibitor U73122 (1μM), its inactive analogue U73433 (1μM) or the scavenger trolox (250μM) was applied during reperfusion. The data presented describe the mean values ± s.e.m. in arbitrary units of fluorescence intensity. They were taken from at least four different experiments. In the absence of these substances the fura 2-ratio significantly increased during reoxygenation from an end-anoxic value of 1.28±0.03 to 1.40±0.03 after 40 min (n=124 cells; p<0.05 vs end-anoxia). Emptying the calcium store by inhibiting its re-entry by THG reduced in the presence of extracelluar calcium after 40 min the calcium rise versus control (Fura-2 ratio: 1.26±0.02; n=120 cells; p<0.05 vs end-anoxic). This suggests an involvement of ER during reoxygenation. Inhibition of the IP3-mediated calcium release by XeC abolished the additional Cai rise during reperfusion (1.26±0.03; n=93 cells; p<0.05 vs without XeC). This indicates that the IP3-receptor dependent release is involved in the reperfusion-induced rise in Cai. Application of the phospholipase C inhibitor U73122 also reduced the Cai rise (1.29±0.03; n=112 cells; p<0.05 vs without U73122) whereas its inactive analogue (1.43±0.02; n=123 cells; ns vs without U73433) had no effect. Trolox, a vitamin E analogue, prevented the additional Cai rise (1.30±0.03; n=98 cells: p<0.05 vs without Trolox). In conclusion, the increase in cytosolic calcium during reoxygenation is due to an influx of extracellular calcium. Additionally, an IP3 dependent calcium release from the ER occurs. The activation of the IP3-release-channel is mediated via activation of phospholipase C. This activation is probably caused by the generation of ROS during reoxygenation.