The selective 5-HT3 receptor agonist 2-methyl-serotonin (2-Me-5-HT) mimicked the antinociceptive activity of 5-HT when intrathecally administered to rats. Two hundred micrograms (i.t.) doses of these agonists produced similar increases in tail flick latency. However, equal doses of 2-Me-5-HT and 5-HT doubled and tripled, respectively, the mean response latency as measured by the hot plate test. The potent and selective 5-HT3 receptor antagonists ICS 205-930 (3-tropanyl-indole-3-carboxylate) and MDL 72222 (3-tropanyl-3,5-dichlorobenzoate) antagonized the antinociceptive effects of both 5-HT and 2-Me-5-HT. However, there were differences in the efficacy of these antagonists. Thus, intrathecal pretreatment with ICS 205-930 (0.05 micrograms) or MDL 72222 (0.1 micrograms) blocked the antinociceptive effects of 5-HT (200 micrograms, i.t.) as measured by the tail flick test, however, higher doses (0.1 and 1.0 micrograms, respectively) were required in the hot plate test. Pretreatment with ICS 205-930 (0.1 microgram) or MDL 72222 (0.1 microgram) blocked the effects of 2-Me-5-HT (200 micrograms, i.t.) in both analgesiometric tests. It is concluded that 5-HT3 receptors are intimately involved in the modulation of spinal nociceptive responses.