Résumé : The regulation of nociceptive processing by 5-HT at the spinal level is intricate since the neurotransmitter has been implicated in both pro and antinociception. The aim of our study was to investigate, according to the nature of the noxious stimulus, how the blockade of spinal 5-HT(1A) receptors could influence the antinociceptive actions of exogenous 5-HT as well as two analgesics involving endogenous 5-HT, paracetamol and venlafaxine. Rats were submitted either to the formalin test (tonic pain) or the paw pressure test (acute pain). WAY-100635 (40 microg/rat, i.t.), a selective 5-HT(1A) receptor antagonist, had no intrinsic action in either test. However, in the formalin test, it blocked the antinociceptive action of 5-HT (50 microg/rat, i.t.) and paracetamol (300 mg/kg, i.v.) in both phases of biting/licking behaviour and that of venlafaxine (2.5 mg/kg, s.c.) in the late phase only. In the paw pressure test, the combination of sub-effective doses of 5-HT (0.01 microg/rat, i.t.), paracetamol (50 mg/kg, i.v.) or venlafaxine (20 mg/kg, s.c.) with WAY-100635 led to a significant antinociceptive effect, which seems to depend on the reinforcement of the activity of inhibitory GABAergic interneurones. In conclusion, both direct stimulation of the spinal 5-HT(1A) receptors by 5-HT, and indirect stimulation using paracetamol or venlafaxine can differently influence pain transmission. We propose that the nature of the applied nociceptive stimulus would be responsible for the dual effect of the 5-HT(1A) receptors rather than the hyperalgesic state or the supraspinal integration of the pain message.