Following the solution of the structure of bovine rhodopsin by X-ray crystallography, it has been possible to build an improved homology model of the M(1) muscarinic acetylcholine receptor. This has been used to interpret the outcome of an extensive series of scanning and point mutagenesis studies on the transmembrane domain of the receptor. Potential intramolecular interactions enhancing the stability of the protein fold have been identified. The residues contributing to the binding site for the antagonist, N-methylscopolamine, and the agonist, acetylcholine have been mapped. The positively charged headgroups of these ligands appear to bind in a charge-stabilized aromatic cage formed by amino acid side chains in transmembrane (TM) helices 3, 6, and 7, while residues in TM 4 may participate in a peripheral docking site. Closure of the cage around the headgroup of acetylcholine may help to transduce binding energy into receptor activation, possibly disrupting a set of Van der Waals interactions between a set of residues underlying the binding site which help to constrain the receptor to the inactive state, in the absence of agonist. This may trigger the reorganization of a hydrogen bonding network between highly conserved residues in the core of the receptor, whose integrity is crucial for activation.