Cholinergic modulation of [3H]acetylcholine release evoked by electrical stimulation was studied in the rat major pelvic ganglion, which was prelabeled with [3H]choline. Acetylcholine (ACh) release was independent of the frequency of stimulation; 0.3 Hz produced the same volley output as 10 Hz. Tetrodotoxin (1 microM) or omission of Ca2+ from the medium abolished ACh release. The M1 receptor agonist (4-hydroxy-2-butynyl)-1-trimethylammonium m-chlorocarbanilate chloride (McN-A 343, 50 microM) increased release (by 136%), whereas the M2 muscarinic agonist oxotremorine (1 microM) decreased ACh release (by 22%). The muscarinic antagonists, atropine (1 microM) or pirenzepine (M1 selective, 1 microM), did not change ACh release. However, pirenzepine (1 microM) blocked the facilitatory effect of McN-A 343, and atropine (1 microM) blocked the inhibitory effect of oxotremorine. The cholinesterase inhibitor physostigmine (1-5 microM), the nicotinic agonist 1,1-dimethyl-4-phenylpiperazinium (DMPP, 10 microM), and the nicotinic antagonist D-tubocurarine (50 microM) did not change ACh release. 4-Aminopyridine, a K+ channel blocker, significantly increased the release (by 146%). Seven days after decentralization of the major pelvic ganglion, the evoked release of ACh was abolished. It is concluded that release of ACh occurs from the preganglionic nerve terminals rather than from the cholinergic cell bodies and is not modulated by actions of endogenous ACh on either muscarinic or nicotinic autoreceptors. These data confirm and extend previous electrophysiological findings indicating that synapses in the major pelvic ganglion have primarily a relay function.