BACKGROUND Intrathecal cholinesterase inhibitors have been shown to have an antinociceptive effect which is mediated through the spinal cholinergic receptors, mainly muscarinic receptor. Spinal nicotinic receptor also has been involved in the control of nociception. Authors characterized the respective role of muscarinic or nicotinic receptor for the antinociception of cholinesterase inhibitors and further determined the antinociceptive potency of them. METHODS Rats were prepared with intrathecal catheters. Formalin-induced flinching response was regarded as a nociceptive behavior. RESULTS Intrathecal neostigmine, physostigmine and edrophonium produced a dose-dependent suppression of flinching in both phases. Atropine and the M1 selective antagonist attenuated the effect of them, while the M2 selective antagonist did not affect. M3, M4 selective, and nicotinic receptor antagonists reversed the antinociception induced by edrophonium, but by neither neostigmine nor physostigmine. The ordering of potency was neostigmine > physostigmine > > edrophonium. CONCLUSIONS These data indicate that the nicotinic receptor may be involved, at least in part, in the antinociceptive action of cholinesterase inhibitor at the spinal level, and M1 receptor subtype may be a common pharmacologic site of action. Moreover, neostigmine is more potent than physostigmine and edrophonium.