Hyperpolarization of arterial smooth muscle by acetylcholine is considered to be produced by the release of an unidentified chemical substance, an endothelium-derived hyperpolarizing factor (EDHF). Several chemicals have been proposed as the candidate for EDHF. However, none of them fulfil completely the nature and property of EDHF. Ultrastructural observation with electron microscope reveals that in some arteries, gap junctions are formed between endothelial and smooth muscle cells. In small arterioles, injection of gap junction permeable dyes into an endothelial cell results in a distribution of the dye to surrounding cells including smooth muscle cells. These observations allow the speculation that myoendothelial gap junctions may have a functional significance. Simultaneous measurement of the electrical responses in both endothelial and smooth muscle cells using the double patch clamp method demonstrates that these two cell types are indeed electrically coupled, indicating that they behave as a functional syncytium. The EDHF-induced hyperpolarization is produced by an activation of Ca2⁢-sensitive K⁢-channels that are inhibited by charybdotoxin and apamin. Agonists that release EDHF increase [Ca2⁢]i in endothelial cells but not in smooth muscle cells. Inhibition of gap junctions with chemical agents abolishes the agonist- induced hyperpolarization in smooth muscle cells but not in endothelial cells. All these observations can be explained if EDHF is an electrotonic signal propagating from endothelium to smooth muscle cells through gap junctions.