Résumé : Antagonist binding to rat pancreatic muscarinic receptors was relatively slow at 25 degrees (tracer dissociation half-life, 50 to 60 min). We, therefore, chose this system to investigate the errors induced by nonequilibrium incubations on the estimates of receptor capacity and selectivity, in binding studies. We took advantage of the fact that muscarinic antagonists recognize only one receptor subtype in rat pancreatic homogenates and that association and dissociation kinetics conform to the law of mass action to analyze quantitatively the binding kinetics of [3H]N-methylscopolamine and of several unlabeled progressive to these receptors. We observed no correlation between the affinities of drugs for muscarinic receptors and their dissociation rate constants. As a result, the apparent receptor specificity (based on relative affinities for different antagonists) varied markedly with the incubation period. We, therefore, strongly recommend that in general competition curves established for receptor classification should be compared at different incubation periods to ensure that equilibrium is attained. The association rate constants of muscarinic antagonists for rat pancreas receptors were remarkably low, when compared with other ligand-receptor systems. This suggests that the antagonist-pancreatic muscarinic receptor association reaction included a rate-limiting conformational change of the drug-receptor complex. This isomerization step was not directly detectable in our kinetic studies, due to the very low affinity and rapid dissociation rate of the initial nonisomerized complex.