The ability of nicotine to induce antinociception after subcortical administration was investigated in the rat. Adult male Sprague-Dawley rats were implanted unilaterally with guide cannulas aimed at the pedunculopontine tegmental nucleus of the mesopontine tegmentum. After 1 week, nicotine was injected in 0.5 microliter of 0.2 M pH 7.4 phosphate buffer. Antinociception was assessed using the 52 degrees C hot-plate test and the tail-flick method; for the most part, the results in the hot-plate test parallelled those in the tail-flick test. Nicotine inhibited nociceptive responses at a median effective antinociceptive dose (A5O) of 1.6 nmol in the hot-plate test and 3.4 nmol in the tail-flick test. Mecamylamine, 0.8 nmol coadministered with nicotine, antagonized nicotine antinociception as evidenced by 5- to 8-fold increases in the nicotine A5O. Nicotine antinociception was also antagonized by coadministrations of either 0.8 nmol of (-)-scopolamine or 0.4 nmol of the M1 antagonist pirenzepine by over 12-fold in the hot-plate test and 5-fold in the tail-flick test. The M2 antagonist methoctramine had antinociceptive effects of its own when injected into the mesopontine tegmentum at a dose of 0.1 nmol; when coinjected with nicotine, the effects of the methoctramine-nicotine combination appeared to be additive. One hour preinjection into the mesopontine tegmentum with 13.5 nmol of (-)-vesamicol, an agent which interferes with acetylcholine storage and/or release, markedly inhibited nicotine antinociception; only a 24% antinociceptive response could be elicited by nicotine in the hot-plate test whereas the nicotine A5O was increased 3-fold in the tail-flick test. Pretreatment with the inactive isomer (+)-vesamicol had no effect. In other experiments, mesopontine tegmental injection of (+)-cis-dioxolane, a high affinity muscarinic cholinergic agonist, elicited strong antinociceptive responses which were potently antagonized by coadministration with 0.5 nmol of pirenzepine but not by 0.8 nmol of mecamylamine. The data indicate that nicotine-induced antinociception may depend upon intact neurotransmission at M1 sites in addition to nicotinic sites within the mesopontine tegmentum.