Muscarinic receptors in sarcolemmal membranes, digitonin-solubilized extracts, and purified preparations from porcine atria have revealed a shortfall in the apparent capacity for N-[3H]methylscopolamine, which was only about 75% of that for [3H]quinuclidinylbenzilate. Since binding at near-saturating concentrations of [3H]quinuclidinylbenzilate was inhibited fully at comparatively low concentrations of unlabeled N-methylscopolamine, the data are inconsistent with the notion that [3H]quinuclidinylbenzilate binds selectively to a subclass of distinct, non-interconverting, and mutually independent sites. The discrepancy is resolved by adjusting the specific activity of N-[3H]methylscopolamine to account for unlabeled scopolamine that was identified in some batches of the radioligand. Also, there was no shortfall in capacity when N-[3H]methylscopolamine was devoid of scopolamine, and the predicted effect was obtained when pure N-[3H]methylscopolamine was supplemented with known amounts of scopolamine. A small discrepancy in the levels of scopolamine estimated pharmacologically and by mass spectrometry can be attributed largely to a difference in the efficiency of ionization between scopolamine and N-methylscopolamine. Different capacities for different radioligands are not uncommon with muscarinic and other G protein-coupled receptors, and in some cases the effect may have been due wholly or in part to an unlabeled impurity. Binding data can be mechanistically ambiguous, particularly when acquired only at graded concentrations of the radioligand. The predicted effects of an unlabeled impurity mimic or resemble those of alternative scenarios such as sequestration behind a hydrophobic barrier, a nucleotide-regulated interconversion from one state of affinity to another, and cooperativity between interacting sites.