The most widely used smooth muscle preparations for neurokinin bioassays have been critically analyzed in order to determine whether neurokinins act directly or by the intermediary of other natural agents. Indeed, part of the contraction of the GPI in response to neurokinins appears to be mediated by acetylcholine and possibly prostaglandins. Active metabolites of the arachidonic acid cascade also intervene in the response of the HUB. Neurokinins produce relaxation of the DCA by stimulating the release of a vascular smooth muscle relaxing factor from the endothelium. In the other preparations (the RD, the RPA without endothelium and the RPV) neurokinins may act directly on the smooth muscle fibers. Neurokinins produce their biological effects by activating specific receptors. Three different receptor types, one for each mammalian neurokinin, have been identified by using four groups of natural peptide sequences and some selective agonists. The receptor for SP is particularly sensitive to SP and physalaemin and shows higher affinity for the whole natural peptides (SP, NKA) than for their C-terminal fragments. The receptor for neurokinin A is highly sensitive to NKA and eledoisin: it shows high affinity for heptapeptide fragments such as NKA4-10 and SP5-11. The receptor for NKB is sensitive to NKB and kassinin more than to the other natural peptides and their fragments. The natural peptides show however little selectivity. Synthetic analogues active on a single receptor type (selective agonists) have been used to find out whether the responses of the isolated organs are due to the activation of one or more than one receptor. It has been found that the GPI, the RD and the HUB contain all three or at least two receptors, while the DCA has only the NK1, the RPA has only the NK2 and the RPV only the NK3 type. Binding sites specific for each neurokinin have been identified in brain and peripheral organs with accurate biochemical assays, using labeled neurokinins. Competitive displacement assays have been performed with a variety of neurokinin-related peptides, and their Ki have been determined. By plotting Ki values against the ED50, estimated from biological assays, positive significant correlations have been found for the monoreceptor (DCA, RPA, RPV) but not for the multiple receptor systems (GPI, RD, HUB). This suggests that pharmacological receptors may be identical with the recognition sites which bind the labeled neurokinins. The availability of monoreceptor systems and of selective agonists opens the way for the identification of potential antagonists and accurate estimation of their affinities.(ABSTRACT TRUNCATED AT 250 WORDS)