1. A range of muscarinic acetylcholine (mACh) receptor agonists (methacholine (MCh), oxotremorine-M (OXO-M), oxotremorine (OXO), arecoline (AREC), bethanechol (BETH), pilocarpine (PILO)) have been investigated with respect to their binding to, and activation of, M(2) and M(4) mACh receptors, recombinantly expressed in Chinese hamster ovary cells, to explore the possibility that these agonists may differentially affect mACh receptor-G(i/o) and -G(s) coupling. 2. M(2)/M(4) mACh receptor coupling to the adenylyl cyclase/cyclic AMP signalling pathway has been explored in intact cells. G(i/o)-mediated negative coupling to adenylyl cyclase was explored functionally by assessing the ability of the mACh receptor agonists to inhibit forskolin-stimulated enzymic activity. Following pertussis toxin treatment (100 ng ml(-1), 18-20 h) to inactivate G(i/o) proteins, each agonist caused a G(s)-mediated enhancement of forskolin-stimulated adenylyl cyclase activity. 3. At both M(2) and M(4) mACh receptors, all agonists tested were more potent in mediating G(i/o)- versus G(s)-coupled responses. This difference (determined as the pIC(50) (G(i/o) coupling) minus pEC(50) (G(s) coupling) value) was greatest for AREC (65-75-fold) and least for BETH and PILO (</=10-fold). 4. Using apparent binding affinities (pK(B)), and potency (EC(50)/IC(50)) and responsiveness (E(max)/I(max)) estimates, relative efficacy (e(rel)) values for each agonist with respect to M(2) and M(4) mACh receptor coupling to G(i/o)- and G(s)-mediated signalling were also calculated. While the e(rel) values obtained for MCh and OXO-M in CHO-m2 cells were similar, OXO-M behaved as a 'super-agonist' at the M(4) mACh receptor giving greater e(rel) values for both G(i/o) and G(s) coupling relative to MCh. 5. The experimental data indicate that while interesting differences between agonists with respect to M(2)/M(4) mACh receptor activation and receptor-G(i/o) and -G(s) coupling can be discerned, no clear examples of agonist trafficking of signal have emerged.