The relative affinities of various muscarinic drugs in the antagonist ([3H]N-methyl scopolamine ([3H]NMS)) and agonist ([3H]Oxotremorine-m ([3H]OXO-M)) binding assays using a mixture of tissues containing M1-M4 receptor subtypes have been determined. [3H]NMS bound with high affinity (Kd = 25 +/- 5.9 pM; n = 3) and to a high density Bmax = 11.8 +/- 0.025 nmol/g wet weight) of muscarinic receptors. [3H]OXO-M appeared to bind to two binding sites with differing affinities (Kd1 = 2.5 +/- 0.1 nM; Kd2 = 9.0 +/- 4.9 microM; n = 4) and to a different population of binding sites (Bmax1 = 5.0 +/- 0.26 nmol/g wet weight; Bmax2 = 130 +/- 60 nmol/g wet weight). Well known antagonists exhibited high affinity for [3H]NMS binding but a lower affinity for [3H]OXO-M binding. The opposite was true for acetylcholine and other known agonists. However, pilocarpine and McN-A-343 had similar affinities for sites labeled by both radioligands. Using the ratios of antagonist-to-agonist binding affinities, it was possible to group compounds into apparently distinct full agonist (ratios of 180-665; e.g. carbachol, muscarine, OXO-M, OXO-S and arecoline), partial agonist (ratios of 14-132; e.g. McN-A-343, pilocarpine, aceclidine, bethanechol, OXA-22 and acetylcholine) and antagonist (ratios of 0.22-1.9; e.g. atropine, NMS, pirenzepine, methoctramine, 4-DAMP and p-fluorohexahydrosialo-difenidol) classes. These data suggest that the NMS/OXO-M affinity ratios using a mixture of M1-M4 muscarinic receptors may be a useful way to screen and group a large number of compounds into apparent agonist, partial agonist, and antagonist classes of cholinergic agents.