In an attempt to determine if the selectivity of secoverine observed in vivo and in isolated tissues might be due to selective association with muscarinic receptor subtypes, we analyzed the binding of secoverine to three different receptors with specific radioligands: rat cardiac receptors (M2 receptors with low affinity for atropine), and rat cerebral cortical M1 receptors and M2 receptors with high affinity for atropine. At concentrations up to 10(-6) M, secoverine interaction with muscarinic receptors was competitive and of high affinity (Ki 4.10(-9) M) for cardiac and brain receptors. A detailed analysis using, in addition to [3H]N-methylscopolamine, the agonist [3H]oxotremorine-M (selective for high affinity binding sites at cardiac receptors) and the M1-selective antagonist [3H]pirenzepine at brain receptors, revealed identical affinities towards both receptor types, making it unlikely that secoverine distinguished the different muscarinic receptor subtypes. At concentrations between 10(-6) and 10(-3) M, secoverine interaction with an additional receptor site resulted in profound changes of tracer kinetics, suggesting the formation of a ternary complex (secoverine-radioligand-muscarinic receptor). The potency of secoverine for provoking this allosteric interaction was both tracer- and tissue-dependent. It is concluded that secoverine does not differentiate between M1, brain M2 and cardiac M2 receptors or between cardiac receptors with high, low and very low affinity for agonists. At very high concentrations secoverine recognized an allosteric site on the muscarinic receptors and reduced the dissociation rates of the 3H-ligands.