The binding characteristics of N[3H]methylscopolamine ([3H]NMS) to slide-mounted tissue sections were studied using quantitative autoradiography. Binding of [3H]NMS was saturable, reversible and of high affinity (Kd = 0.26 nM). The inhibition of [3H]NMS binding produced by several muscarinic agonists and antagonists was analyzed in 29 discrete brain regions by constructing complete displacement curves. Comparison of IC50 values obtained both biochemically and by autoradiography demonstrated a very close agreement, supporting the validity of the autoradiographic approach. The competition curves for the agonists carbachol, oxotremorine and 2-ethyl-8-methyl-2,8-diazaspiro-[4,5]-decan-1,3-dion-h ydrobromide (RS 86) fitted to a two-site model, with comparable affinity values from region to region, although different proportions of high- and low-affinity sites were seen in the different areas studied. The distribution of high- and low-affinity sites was similar for the three agonists. Atropine showed monophasic curves presenting similar affinities in all regions studied. In contrast, pirenzepine differentiated between high- and low-affinity sites which showed a distribution opposite to that observed for the agonists. Gallamine, a ligand for a putative regulatory site in the muscarinic receptor, inhibited [3H]NMS binding in a biphasic manner. The calculated IC50 values for the gallamine high- and low-affinity sites did not vary from region to region and the distribution of these sites correlated well with that observed for the agonists. High-affinity pirenzepine sites (also called M1 sites) were localized mainly in forebrain areas, such as striatum, hippocampus and cortex, and their regional distribution correlated with that of the low-affinity sites for the agonists and gallamine. On the other hand, low-affinity sites for pirenzepine (named M2 sites) were mainly found in the brainstem and parts of the thalamus. A good correlation was found between pirenzepine low-affinity sites and agonist and gallamine high-affinity sites. The significance of these findings is discussed in relation to the known and possible effects of selective M1 and M2 centrally acting agents.