1 In the light of recent attempts to subclassify muscarinic receptors, agonist-antagonist interactions at muscarinic receptors have been re-examined using improved techniques, on the mouse, isolated, lumen-perfused stomach gastric acid assay.2 Using 5-methylfurmethide as the muscarinic agonist, the pKB estimated for atropine was significantly lower on the stomach assay (7.78) than on the guinea-pig trachea (8.93). However pKB values for N-methylatropine, the quaternary ammonium derivative of atropine, at concentrations producing dose-ratios above 20 on the stomach assay (pKB = 9.67), and over the full concentration range studied on the trachea (pKB = 9.69) were not significantly different.3 The deviation from simple competitive behaviour at low dose-ratios with N-methylatropine in the stomach assay is consistent with the effects of a saturable uptake mechanism for quaternary ammonium compounds. 4 The pKB values for pirenzepine on the stomach (6.67) and the trachea (6.87) were not significantly different suggesting that pirenzepine behaves more like N-methylatropine in terms of expressed affinity. 5 We conclude that the oxyntic cell muscarinic receptors are homogeneous with those in the guineapig trachea. An initial exploration suggests that there is a relationship between the lipophilicity (log P) of the antagonists and the degree of apparent underestimation of antagonist affinity in the stomach assay. This supports the hypothesis that the underestimation of antagonist affinity is due to the loss of antagonist into the gastric secretion from the receptor compartment. Apparently, relatively selective inhibition of acid secretion, compared to atropine, could be explained without the need to postulate heterogeneity of muscarinic receptor populations.