We investigated which of the major actions of local anesthetics (i.e., inhibition of phospholipase A2, interaction with Ca++ channels or blockade of receptor) was responsible for the inhibition of acetylcholine-induced desensitization in guinea pig ileal longitudinal muscle. Desensitization was inhibited by amine local anesthetics and related compounds in the order of potency quinacrine > chloroquine > tetracaine > procaine. Potent phospholipase A2 inhibitors, manoalide (1 microM) and p-bromophenacyl bromide (5 microM) had no effect on desensitization. The rank order of interaction of local anesthetics with Ca++ channels did not agree with the potency order of inhibition of desensitization. These data indicated that local anesthetics did not inhibit desensitization through their inhibition of phospholipase A2 or their interaction with Ca++ channels. Quinacrine, chloroquine, tetracaine and procaine inhibited [3H]N-methylscopolamine binding to solubilized membrane with pKi values of 7.03 +/- 0.10, 6.59 +/- 0.02, 5.40 +/- 0.10 and 5.03 +/- 0.04 and reduced receptor occupancy by agonist from 99.0% (without inhibitor) to 96.8%, 95.1%, 89.4% and 49.8%, respectively, under the conditions where each drug induced half-maximum inhibition of desensitization, indicating that they (except for procaine) did not effectively block muscarinic receptors. However, the combined dose-ratio test showed that some of these drugs (quinacrine and chloroquine) interacted noncompetitively at muscarinic receptors. Therefore, these drugs could have bound to an allosteric site on the receptor, modified agonist-receptor interaction and thus inhibited the pathway specific to the desensitization process.