Both systemically and intrathecally administered cholinergic agonists produce antinociception while cholinergic antagonists decrease pain threshold. The mechanism and the site of action of these substances are not known. In the present study it was hypothesized that systemically administered muscarinic agonists and antagonists modify nociceptive threshold by affecting intraspinal release of acetylcholine (ACh). Catheters were inserted into the femoral vein in rats maintained on isoflurane anaesthesia for administration of oxotremorine (10-300 microg/kg) and atropine (0.1, 10, 5000 microg/kg). Spinal microdialysis probes were placed intraspinally at approximately the C2-C5 spinal level for sampling of acetylcholine and dialysis delivery of atropine (0.1, 1, 10 nM). Additionally, the tail-flick behaviour was tested on conscious rats injected intraperitoneally with saline, atropine (10, 100 and 5000 microg/kg), or subcutaneously with oxotremorine (30, 100, 300 microg/kg). Subcutaneous administration of oxotremorine (30, 100, 300 microg/kg) significantly increased the tail-flick latency. These doses of oxotremorine dose-dependently increased the intraspinal release of acetylcholine. Intravenously administered atropine, in a dose that produced hyperalgesia (5000 microg/kg) in the tail-flick test, significantly decreased the intraspinal release of acetylcholine. Our results suggest an association between pain threshold and acetylcholine release in spinal cord. It is also suggested that an approximately 30% increase in basal ACh release produces antinociception and that a 30% decrease in basal release produces hyperalgesia.