This study evaluated the ability of two new, selective antagonists of the gamma-aminobutyric acid(B) (GABA(B)) receptor, phaclofen (PHAC) and 2-hydroxy-saclofen (2-OH-S), to antagonize the increase in tail-flick latency (TFL) and hot-plate latency (HPL) produced by i.t. administered baclofen (BAC) in the rat. The putative GABA(B) receptor antagonist delta-aminovaleric acid (DAVA) was also examined for comparative purposes. Intrathecal(i.t.) pretreatment with increasing doses of PHAC (10-100-mu-g) shifted the dose-effect relationship of i.t. administered BAC progressively to the right in a parallel manner in both the tail-flick (TF) and hot-plate (HP) test. Schild analysis of the data yielded an apparent pA2 value of 7.3 +/- 0.1 and a slope of -0.98 +/- 0.14. By comparison, PHAC did not antagonize the increase in HPL produced by i.t. injection of the serotonin1A agonist, 8-hydroxy-N,N-dipropyl-2-aminotetralin. These observations indicate that PHAC competitively and selectively antagonizes BAC and further suggest that the antinociceptive effects of i.t. administered BAC are mediated by the PHAC-sensitive subtype of the GABA(B) receptor. Intrathecal injection of PHAC alone did not decrease TFL or HPL, suggesting that spinal GABA(B) receptors involved in nociception are not tonically activated. Although i.t. pretreatment with 2-OH-S (10-30-mu-g) also antagonized the antinociceptive effects of i.t. administered BAC, increasing doses of 2-OH-S did not produce progressive, rightward shifts in the dose-effect relationship of BAC. Indeed, i.t. administration of 2-OH-S alone modestly increased TFL, but not HPL in the rat. These observations suggest that 2-OH-S may be a partial agonist at spinal GABA(B) receptors. Finally, i.t. pretreatment with DAVA (10-30-mu-g) did not antagonize BAC-induced antinociception. However, a high dose of DAVA (100-mu-g i.t.) enhanced the increase in TFL produced by BAC, an observation that may reflect its ability to inhibit GABA uptake.