1. We have studied several tachykinin NK2 receptor antagonists, bearing a monocyclic pseudopeptide (MEN 10,508, MEN 10,573, MEN 10,581, MEN 10,612, MEN 10,619 and MEN 10,677), or bicyclic peptide (MEN 10,627, MEN 10,692, MEN 10,771, MEN 10,882 and MEN 10,993) structure, on the human isolated urinary bladder detrusor muscle against neurokinin A as an agonist, and compared their affinities in this preparation with those for NK2 receptors expressed in the rabbit isolated pulmonary artery and hamster isolated trachea. 2. In the human bladder, all the antagonists tested produced a concentration-dependent and competitive antagonism of neurokinin A-mediated contractions: among the cyclic pseudopeptides MEN 10,677 (pKB = 8.0) was the most potent antagonist, while among the bicyclic analogues it was MEN 10,993 (pKB = 8.8). 3. In general, the bicyclic peptide antagonists tested were more potent than the monocyclic pseudopeptide compounds, either in the human urinary bladder or in the rabbit pulmonary artery or hamster trachea, showing a nanomolar affinity for the human NK2 receptor. 4. A highly significant correlation was found between the estimated pKB values of all the antagonists tested in the human urinary bladder and rabbit pulmonary artery (r2 = 0.94, n = 12, P < 0.01), whereas no linear correlation was found between pKB values measured in the human urinary bladder and hamster trachea (r2 = 0.52, n = 12, P > 0.05): these observations provide further pharmacological evidence for receptor homology between the human and rabbit NK2 receptor. 5. The present results point out the class of NK2 receptor antagonists bearing a bicyclic peptide structure, like MEN 10,627, as candidates for testing in pathological conditions, such as bladder hyperactivity, for which preclinical evidence indicates that a therapeutic effect could result from the block of the tachykinin NK2 receptor.