Radioligand binding and quantitative autoradiographic techniques were used to characterize the pharmacological profile and anatomical distribution of N-[3H]methylscopolamine [( 3H] NMS)-labeled sites under assay conditions in which radioligand binding to M1 and M2 muscarinic receptors was blocked by addition of pirenzepine and AF-DX 116 (11[[2-[(diethylamino)methyl]-1-piperidinyl] acetyl]-5,11-dihydro-6H-pyrido[2,3-b] [1,4]benzodiazepine-6-one) to the incubation buffer. Nonlinear regression analysis of saturation data demonstrated that a large proportion of atropine-displaceable [3H]NMS binding persisted in the presence of saturating concentrations of M1 and M2 blockers. The residual [3H]NMS [( 3H]NMSb) sites were widely distributed throughout rat brain and represented the predominant muscarinic receptor population. The autoradiographic distribution of [3H]NMSb sites did not correspond to that of [3H]pirenzepine or [3H]AF-DX 116, indicating that [3H]NMSb labeled non-M1, non-M2 muscarinic sites. Moreover, the pharmacological profile of [3H]NMSb differed from that of [3H]pirenzepine at M1 sites, and was inconsistent with that of M2 receptor binding sites. Although we were unable to pharmacologically distinguish subpopulations of non-M1, non-M2 binding sites, the anatomical distribution of [3H]NMSb sites corresponded to that of the combined mRNA distributions for m3 and m4.