Two muscarinic toxins, MT1 and MT7, were obtained by one-step solid-phase synthesis using the 9-fluorenylmethoxycarbonyl-based method. The synthetic and natural toxins, isolated from the snake venom or recombinantly expressed, display identical physicochemical properties and pharmacological profiles. High protein recovery allowed us to specify the selectivity of these toxins for various muscarinic receptor subtypes. Thus, sMT7 has a selectivity for the M1 receptor that is at least 20,000 times that for the other subtypes. The stability of the toxin-receptor complexes indicates that sMT1 interacts reversibly with the M1 receptor, unlike sMT7, which binds it quasi-irreversibly. The effect of the synthetic toxins on the atropine-induced [3H]N-methylscopolamine (NMS) dissociation confirms that sMT7 targets the allosteric site on the M1 receptor, whereas sMT1 seems interact on the orthosteric one. The great decreases in the binding potencies observed after the R34A modification in sMT1 and sMT7 toxins highlight the functional role of this conserved residue in their interactions with the M1 receptor. Interestingly, after the R34A modification, the sMT7 toxin binds reversibly on the M1 receptor. Furthermore, the potency of sMT7-R34A for the NMS-occupied receptor is lower compared with unmodified toxin, supporting the role of this residue in the allosteric interaction of sMT7. All these results and the different charge distributions observed at the two toxin surfaces of their structure models support the hypothesis that the two toxins recognize the M1 receptor differently.