The phenomenon of endothelium-dependent hyperpolarization and relaxation and the possible existence of a factor – endothelium-derived hyperpolarizing factor (EDHF) – has been widely studied for the past 25 years. The broad consensus view is as follows:- Classical EDHF Pathway; gap junctions or K⁺? If the endothelial synthesis of NO and prostacyclin is blocked using NO-synthase and cyclo-oxygenase inhibitors, ligands like acetylcholine (ACh) and substance P (SP) still hyperpolarize and relax vascular myocytes in an endothelium-dependent manner. These effects can be blocked by a mixture of TRAM-34 + apamin (blockers of endothelial IK_Ca and SK_Ca channels, respectively) but not by iberiotoxin + apamin. This then is the pharmacological signature of the (classical) EDHF pathway. ACh and SP raise endothelial [Ca²⁺], opening IK_Ca and SK_Ca channels and hyperpolarizing the endothelial cells. Controversy still surrounds the mechanism by which the endothelial hyperpolarization is transmitted to the myocytes to trigger vasodilation. The gap junction hypothesis claims that there is no EDHF per se; rather the hyperpolarization is transmitted to the myocytes via myo-endothelial gap junctions. The K⁺ hypothesis proposes that EDHF is K⁺. This ion, effluxing from endothelial cells through opened IK_Ca and SK_Ca channels activates myocyte K_IR channels and Na⁺/K⁺ ATPases; the resulting hyperpolarization produces the observed vasodilation. It is likely that both hypotheses are correct, the importance of each being dependent on the vessel and its degree of tone. Non-Classical Pathways; other hyperpolarizing factors The widespread use of bradykinin (especially in the coronary circulation) to activate the classical EDHF pathway described above, inadvertently created confusion in the literature and amongst EDHF workers. In addition to activating the classical pathway, bradykinin also releases epoxyeicosatrienoic acids (EETs; most likely both 11,12 EET and 14,15 EET) from the endothelium. These diffuse to the myocytes and hyperpolarize these cells by activating myocyte BK_Ca channels. Thus, in early experiments on the EDHF pathway, the widespread use of the blocking mixture charybdotoxin (TRAM-34 was not available) + apamin was certainly effective in abolishing ‘EDHF’ responses mediated by bradykinin. However, it was not then realised that charybdotoxin not only blocked endothelial cell IK_Ca channels but also myocyte BK_Ca channels and thus both the classical EDHF pathway and the EETs pathway were inhibited. What indeed is EDHF? Based on the original observations, the classical pathway can best be described as that endothelium-dependent vasodilator pathway that is blocked by TRAM-34 + apamin. For those who feel that myo-endothelial gap junctions are the key to the resulting vasodilator response, no actual mediator is necessary. However, when vascular tone is moderate, considerable evidence identifies EDHF as endothelium-derived K⁺. Indeed the involvement of K⁺, locally released from neurones and skeletal muscle cells to increase blood flow within the brain and skeletal muscle, respectively, is well-established. EETs, prostacyclin (and sometimes even NO) are each liberated from the vascular endothelium and each can hyperpolarize the adjacent myocytes. They and other endothelium-derived substances could thus be described as EDHFs. However, the involvement of BK_Ca or of K_ATP in their actions means that it is inappropriate to give them the descriptor ’EDHF’.