[en] The in vivo quantification of myocardial muscarinic receptors has been obtained in six closed-chest dogs by using positron emission tomography. The dogs were injected with a trace amount of 11C-labeled methylquinuclidinyl benzilate (MQNB), a nonmetabolized antagonist of the muscarinic receptor. This was followed 30 minutes later by an injection of an excess of unlabeled MQNB (displacement experiment). Two additional injections of unlabeled MQNB with [11C]MQNB and without [11C]MQNB (second displacement experiment) were administered after 70 and 120 minutes, respectively. This protocol allowed a separate evaluation of the quantity of available receptors (B'max) as well as the association and dissociation rate constants (k+1 and k-1) in each dog. The parameters were calculated by using a nonlinear mathematical model in regions of interest over the left ventricle and the interventricular septum. The average value of B'max was 42 +/- 11 pmol/ml tissue, the rate constants k+1, k-1, and Kd were 0.6 +/- 0.1 ml.pmol-1.min-1, 0.27 +/- 0.03 ml.pmol-1.min-1, and 0.49 +/- 0.14 pmol.ml-1, respectively, taking into account the MQNB reaction volume estimated to 0.15 ml/ml tissue. Although [11C]MQNB binding would appear irreversible, our findings indicate that the association of the antagonist is very rapid and that the dissociation is far from negligible. The dissociated ligand, however, has a high probability of rebinding to a free receptor site instead of escaping into the microcirculation. We deduce that the positron emission tomographic images obtained after injecting a trace amount of [11C]MQNB are more representative of blood flow than of receptor density or affinity. We also suggest a simplified protocol consisting of a tracer injection of [11C]MQNB and a second injection of an excess of cold MQNB, which is sufficient to measure B'max and Kd in humans