This study examined whether the antinociceptive activity of THIP (4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol), a GABA agonist, is mediated through an action exerted at the level of the spinal cord. Intrathecal injection of doses of THIP devoid of motor effects (1-2 micrograms) did not increase tail flick or hot plate latencies in the rat. Although hot plate latency was transiently increased by intrathecal injection of 5 micrograms THIP, slight motor impairment was observed at this dose. Higher doses of THIP (15-50 micrograms) produced flaccidity of the hindlimbs. Intrathecal injection of low doses of muscimol (0.25 microgram) that did not produce motor impairment increased tail flick, but not hot plate, latencies. Baclofen produced motor incoordination following intrathecal injection of 10 micrograms; however, intrathecal injection of 1 microgram significantly increased both tail flick and hot plate latencies without attendant motor effects. Thus, baclofen was the only compound in which the antinociceptive effect was clearly distinguished from the motor effect. These results additionally indicate that the spinal cord does not mediate the antinociceptive activity of THIP.