To examine the role of the 5-hydroxytryptamine(1B) (5-HT1B) and 5-HT3 receptor subtypes in the analgesia produced by 5-HT (serotonin) agonists, we assessed the effect of antisense oligodeoxynucleotides (AODNs) designed to "knock down" the number of these receptor subtypes on analgesia produced by intrathecal (i.t.) 5-HT, the 5-HT1B receptor agonist, 7-trifluoromethyl-4-(4-methyl-1-piperazinyl)-pyrrolo[1,2-a]quinoxaline maleate (CGS-12066A), and the 5-HT3 receptor agonist, 2-methyl-5-HT. Groups of mice (n = 17-20) were injected i.t. on days 1, 3, and 5 with one of the AODNs, a mismatch oligo, or saline. On day 6, all mice were injected i.t. with 70.5 nmol of 5-HT, 44.4 nmol of CGS-12066A, or 49 nmol of 2-methyl-5-HT by lumbar puncture. Following testing, spinal cords were rapidly removed and prepared for receptor binding assays. Treatment with AODN for 5-HT1B receptors produced a 70% reduction in ligand binding to this receptor subtype. After treatment with AODN for 5-HT3 receptors, ligand binding to this receptor subtype was undetectable. In mice tested with i.t. 5-HT, tail-flick analgesia was attenuated only in mice treated with the 5-HT3 receptor AODN. Mice treated with the AODN designed to knock down 5-HT(1B) receptors or with its mismatch oligo were not significantly different from controls. In mice tested with i.t. administration of CGS-12066A, none of the oligo treatments produced a significant attenuation of analgesia. In mice tested with i.t. administration of 2-methyl-5-HT, only 5-HT3 receptor AODN attenuated analgesia. Thus, 5-HT and 2-methyl-5-HT analgesia are mediated by the 5-HT3 receptor subtype. However, spinal CGS-12066A analgesia appears not to be mediated by either the 5-HT1B or the 5-HT3 receptor subtypes.