Olanzapine and clozapine produce robust increases in hippocampal acetylcholine release during acetylcholinesterase inhibition, while other antipsychotics, including thioridazine, have only small effects. Since thioridazine binds with similar high affinities to muscarinic receptors as olanzapine and clozapine, muscarinic autoreceptor blockade was ruled out as a primary mechanism [Neuropsychopharmacology 26 (2002) 583]. This study compared in vitro binding affinities and functional activities of olanzapine, clozapine, thioridazine, ziprasidone, risperidone, chlorpromazine and scopolamine at muscarinic M2 receptors with their in vivo potencies to increase acetylcholine release in the rat hippocampus. We found that scopolamine, olanzapine and clozapine, but also high doses of thioridazine and chlorpromazine, markedly increase acetylcholine release. The reduced in vivo potencies of thioridazine and chlorpromazine are consistent with their significantly weaker functional antagonist activity at human muscarinic M2 receptors, while thioridazine's reduced binding affinity for rat muscarinic M2 receptors and lower brain exposure, may further contribute to its weak in vivo potency compared to olanzapine. The excellent correlation between in vitro antagonist activities of antipsychotics at muscarinic M2 receptors and their in vivo potencies to increase acetylcholine release, suggests that olanzapine, clozapine, as well as thioridazine and chlorpromazine, increase acetylcholine release via blockade of terminal muscarinic M2 autoreceptors.