1 The effect of muscarinic antagonists considered to be selective for M1 receptors (pirenzepine) and for M2 receptors (gallamine) were studied on bronchoconstriction and bradycardia elicited by stimulation of the vagal nerves and by i.v. acetylcholine (ACh) in anaesthetized rabbits. 2 Pirenzepine was equipotent as an antagonist of ACh-induced responses at postjunctional muscarinic receptors in the heart, lung and blood vessels, whereas gallamine was at least ten times less potent at pulmonary and vascular muscarinic receptors. Thus, gallamine never caused complete inhibition of bronchoconstrictor or hypotensive responses to i.v. ACh, whereas doses of pirenzepine in excess of 1 pmol kg1-abolished all muscarinic responses. 3 In the lung, both antagonists inhibited bronchoconstriction caused by vagal stimulation and ACh-induced bronchoconstriction to the same extent (pirenzepine, mean ED50 65 + 22 and 130 + 28nmolkg-1 respectively; gallamine, ED50> 10,OOOnmolkg-1 for both responses).Enhancement of vagally-induced bronchoconstriction was never observed. 4 In the heart, however, both pirenzepine and gallamine were ten times less potent as antagonists of vagally-induced bradycardia than of ACh-induced bradycardia. This differential blockade was unaltered by propranolol (1 mg kg-1) pretreatment. 5 It is concluded that there is no evidence for M1 or M2 muscarinic receptors in the pulmonary innervation of the rabbit and the potency of the antagonists in abolishing vagally-induced bronchoconstriction was consistent with blockade of M3 muscarinic receptors on airway smooth muscle. 6 The results suggest that M2 muscarinic receptors may exert an inhibitory effect on transmission in the parasympathetic nerves innervating the heart in the rabbit. Blockade of such neuronal receptors would increase transmitter output to the atrial cells and explain the low potency of both antagonists in abolishing vagally-induced bradycardia in the rabbit.