To explain the central hypotensive action of clonidine- and rilmenidine-type drugs, a recent hypothesis suggested the involvement of imidazoline preferring receptors (IPR) insensitive to catecholamines. Binding studies performed on neuronal membranes prepared from the human Nucleus Reticularis Lateralis area (NRL) showed that rilmenidine was twice as selective as clonidine for the medullary IPRs. Voltammetric experiments revealed that these substances preferentially inhibited the activity of catecholaminergic neurones within the NRL, privileged site for their hypotensive action, rather than the activity of the neurones of the Locus Coeruleus (LC) where these drugs induce their sedative effect. The mechanism of the inhibitory action of these compounds upon the neurones of the NRL definitely involved IPRs when the inhibition of the LC neurones involved classical alpha-2-adrenoceptors. These functional studies also showed that rilmenidine exhibited a selectivity twice as important as clonidine, the reference substance. Poly- and monoclonal anticlonidine antibodies cross-reacting only with imidazolines allowed to detect and immunoreactive substance in the human serum. The amounts of that substance thus measured were higher in some hypertensive patients than in normotensive subjects. The identity of that immunoreactive substance with "endazoline", the endogenous ligand of the IPRs, is presently analysed.