Muscarinic acetylcholine receptors are G protein-coupled receptors (GPCRs) that regulate many important functions of the central and peripheral nervous system. The M1 and M4 muscarinic acetylcholine receptors (M1/M4 receptor) have emerged as attractive drug targets for treatment of psychosis and cognitive deficits. However, muscarinic drug development has been hampered by a high degree of sequence conservation at the orthosteric acetylcholine-binding site, making it very difficult to develop subtype selective drugs. In recent years, there has been substantial progress in the discovery of more selective ligands that act at allosteric sites. How allosteric drugs exert their actions at these receptors, however, is not fully understood. To better appreciate the basis of selectivity and allostery at the M4 receptor we combined structural information with mutagenesis studies. We first determined the crystal structure of the M4 receptor bound to the antagonist, tiotropium. Overall, the tiotropium-bound M4 structure shares many similarities with the inactive M2 and M3 receptor structures, however we found some notable differences in the orthosteric ligand-binding site. In an effort to further dissect the molecular mechanism of allostery at the M4 receptor, we utilized our inactive-state M4 structure together with an active-state M4 model (based on the recently solved active-state structure of the homologous M2 muscarinic receptor¹) to rationalize the effects of targeted mutations on the interaction between a positive allosteric modulator and acetylcholine. In doing so, we identified an allosteric “hotwire” that links the allosteric and orthosteric sites and is composed primarily of the interfaces between TMs 2, 3 and 7. Overall, our findings indicate that it is possible to combine crystal structure data with mutagenesis data to uncover new insights into GPCR allosteric modulation.