Hydrogen peroxide (H₂O₂) is a proposed endothelium-derived hyperpolarising factor that is believed to act on iberiotoxin-sensitive calcium-activated potassium channels (BK_ca) located on the smooth muscle. However it has also been demonstrated that relaxation induced by exogenously applied H₂O₂ is blocked by the combination of charybdotoxin and apamin, drugs which block endothelial K_ca channels. The aim of this study was to investigate the role of endothelial K_ca channels in the vasodilating action of H₂O₂. Male Wistar rats (250-300g) were killed by cervical dislocation. Third order mesenteric arteries were mounted on a Mulvany-Halpern small vessel wire myograph for measurement of isometric contraction. All experiments were performed in 300μM L-NAME. All results are expressed in mean ± SEM and statistical analysis was performed using Tukeys Multiple Comparison Test or Students T-test. 100 μM H₂O₂ caused a 78±4% relaxation (n=4) under control conditions. Relaxation was abolished by 100 nM iberiotoxin (2±2% relaxation), and was inhibited by endothelial IK_ca and SK_Ca channel inhibitors TRAM-34 (1μM) and apamin (100nM) (24±8% relaxation, n=4, p<0.001). To investigate the role of myoendothelial gap junctions, 18α-glycyrrhetinic acid (18αGA, 100μM) was applied for 1 hour. Despite blocking relaxation induced by acetylcholine (ACh, control: 39±5%, +18αGA: 6±1%, n=4, p<0.001), there was no inhibitory effect of the drug on H₂O₂ relaxations (+18αGA: 77±6% relaxation, n=4, n.s.). Unexpectedly, in the presence of 18αGA, TRAM-34 and apamin did not inhibit the response to H₂O₂ (75±3% relaxation, n=4, P<0.001). The effect by iberiotoxin (100nM) was insensitive to 18αGA. To compare these findings with a second agent that acts by hyperpolarising smooth muscle, the K_ATP channel opener levcromakalim (2 μM) was studied and similar results to the H₂O₂ findings were obtained. TRAM-34 and apamin significantly inhibited relaxation to levcromakalim (control relaxation: 70±12, + TRAM-34 and Apamin: 13%±3, n=5, P<0.001), but this effect was reversed by 18αGA (71%±5 relaxation, n=5, P<0.001). These results suggest that vasorelaxation to agents acting by hyperpolarising vascular smooth muscle can be prevented by inhibiting endothelial hyperpolarisation via IK_Ca and SK_Ca channels. Thus, coupling of the endothelium to the smooth muscle via myoendothelial gap junctions may not only elicit hyperpolarisation during the EDHF response, but could act to restrict vasorelaxation by agents which act exclusively on the smooth muscle.