Subcutaneous injection of formalin into a paw of mice caused two distinct phases of licking and biting, first phase (1-5 min) and the second phase (7-30 min) after the injection. The muscarinic antagonist atropine (0.1-10 ng, i.t.) and the M(3) receptor antagonist 4-diphenylacetoxy-N-methylpiperidine methiodide (4-DAMP) (0.1-20 ng, i.t.) inhibited the second phase of this response, whereas higher doses of atropine (20-100 ng, i.t.) did not cause inhibition. The M(1) muscarinic receptor antagonist pirenzepine (10-100 ng, i.t.) did not inhibit either the first or the second phase response, but a high dose of pirenzepine (1000 ng, i.t.) tended to inhibit the second phase response. On the other hand, the M(2) muscarinic receptor antagonist 11-¿(2-[(diethylamino)methyl]-1-piperidinyl¿acetyl)-5, 11-dihydro-6H-pyrido(2,3-b)(1,4)benzodiazepine-6-one (AF-DX116; 10-1000 ng, i.t.) had no effect on either the first or the second phase of response. The opioid receptor antagonist naloxone did not affect the 4-DAMP-induced anti-nociceptive response. The i.t. injection of the acetylcholinesterase inhibitor neostigmine (25 ng) significantly inhibited only the second phase. The acetylcholine (ACh) depletor hemicholinium-3 (HC-3) (1 microg, i.t.) completely abolished the 4-DAMP-induced anti-nociceptive response. The ACh content of the spinal cord was significantly increased 14 min after formalin injection. This significant increase in the ACh content was inhibited by pretreatment with 4-DAMP (10 ng, i.t.). These results suggest that endogenous ACh in the spinal cord acts as a transmitter anti-nociception, and that ACh release regulated by presynaptic M(3) muscarinic receptors in the spinal cord is involved in the second phase of nociception induced by formalin.