1 Affinity estimates were obtained for several muscarinic antagonists against carbachol-stimulated [H-3]-inositol phosphates accumulation in Chinese hamster ovary CHO-K1(I) cells stably expressing either human muscarinic M-3 or MS receptor subtypes. The rationale for these studies was to generate a functional antagonist affinity profile for the Mg receptor subtype and compare this with that of the M-3, receptor, in order to identify compounds which discriminate between these two subtypes. 2 The rank order of antagonist apparent affinities (pK(B)) at the muscarinic M-5 receptor was atropine (8.7)tolterodine (8.6)= 4-diphenylacetoxy-N-methylpiperidine (4-DAMP, 8.6)>darifenacin (7.7)greater than or equal to zamifenacin (7.6)>oxybutynin (6.6) =para-fluorohexahydrosiladifenidol (p-F-HHSiD, 6.6) > pirenzepine (6.4) greater than or equal to methoctramine (6.3) = himbacine (6.3) > AQ-RA 741 (6.1). 3 Antagonist apparent affinities for both receptor subtypes compare well with published binding affinity estimates. No antagonist displayed greater selectivity for the muscarinic M-5 subtype over the Mt subtype, but himbacine, AQ-RA 741, p-F-HHSiD, darifenacin and oxybutynin displayed between 9- and 60 fold greater selectivity for the muscarinic Mt over the M-5 subtype, 4 This study highlights the similarity in pharmacological profiles of M-3 and MS receptor subtypes and identifies five antagonists that may represent useful tools for discriminating between these two subtypes. Collectively, these data show that in the absence of a high affinity MS selective antagonist, affinity data for a large range of antagonists is critical to define operationally the MS receptor subtype.