We have recently demonstrated (J Physiol 506 (1998) 459) that the dynamic activation of descending inhibition of the nociceptive response of spinal multireceptive cells occurs in the nucleus reticularis gigantocellularis pars alpha (GiA). In the same paper we have shown that Lamina I dorsal horn cells are responsible for activating this inhibition via a pathway which runs in the contralateral dorsolateral funiculus. The effects of dynamically activating this system by noxious stimulation on behavioural responses to noxious stimuli have not been established. Here we demonstrate the effects of GiA on the behavioural response during application of standardized noxious stimuli. As this system is activated in response to noxious stimulation (J Physiol 506 (1998) 459), it is possible that chronic pain states may also activate GiA. We have therefore investigated this possibility in animals following partial sciatic nerve ligation (an animal model of chronic pain; Pain 43 (1990) 205). Male Wistar rats (280-310 g) were anaesthetized with halothane (0.5-2% in O(2)). Guide cannulae for microinjections were stereotaxically placed above GiA. In one group of animals the sciatic nerve was partially ligated. Animals were allowed to recover for 4-6 days. The responses of each animal during the formalin test (Pain 4 (1977) 161) and the tail flick test (Pain 12 (1982) 229) were recorded on different days. Microinjections (0.5 microl) of either gamma-aminobutyric acid (GABA, 200 mM), D-L homocysteic acid (DLH, 25 mM) or 0.9% saline (as control) into GiA were performed during these tests in a randomized, blind manner. In animals without sciatic nerve ligation, microinjection of GABA to GiA did not significantly affect the animal's response during the tail flick test. However microinjection of DLH significantly increased the latency of tail flick from 6.2 +/- 0.8 to 8.4 +/- 0.5 s for up to 15 min (n = 7, P < 0.01, Mann-Whitney U-test). Microinjection of GABA to GiA increased the behavioural response to formalin between 10 and 20 min post-injection, while microinjection of DLH reduced this response at all time points except 10 min post-injection (n = 8, P < 0.05, Mann-Whitney U-test). In animals with sciatic nerve ligation, microinjections (0.5 microl) of either GABA (200 mM), or saline (as control) into GiA contralateral to the partial sciatic ligation were performed during these tests in a randomized, blind manner. Partial sciatic ligation significantly reduced the behavioural response to contralaterally applied formalin from 15 min post-injection onwards, compared to controls without sciatic nerve ligation. Microinjection of GABA to GiA significantly increased the behavioural response to formalin from 20 to 50 min post-injection. The inactivation of GiA only causes behavioural effects in nociceptive tests of a long enough duration to activate the system (i.e. the formalin test but not the tail flick test). Chemical activation of the system affects both tests. These data strongly support the concept of an important analgesic system which is activated in response to noxious stimulation, and subsequently acts to reduce behavioural responses to noxious stimuli.