We have compared the ability of the new putatively specific 5-HT1B receptor agonist CP-93,129 (3-(1,2,5,6-tetrahydropyrid-4-yl)pyrrolo[3,2-b]pyrid-5-one) and the structurally related mixed 5-HT1A/5-HT1B receptor ligand RU 24969, to influence 5-HT release in brain in vivo, using microdialysis techniques in chloral hydrate-anaesthetised rats. CP-93,129 (3 or 10-mu-M, via the dialysis perfusion medium) caused a concentration-dependent and methiothepin (10-mu-M)-sensitive suppression of ventral hippocampal 5-HT output, The effect of RU 24969 on 5-HT output was dependent on whether or not the 5-HT reuptake blocker citalopram was present in the perfusion medium. Thus, RU 24969 (0.1-mu-M) induced a decrease, or an increase followed by a decrease (1-mu-M), in 5-HT output in the absence of citalopram, but monotonically decreased (1-mu-M) 5-HT release when citalopram (1-mu-M) was present. CP-93,129 decreased dialysate 5-HT in either condition. Our findings are consistent with the characterisation of CP-93,129 as a 5-HT1B receptor agonist, and may thus represent in vivo support for 5-HT1B autoreceptor-mediated feedback control of 5-HT release in the rat brain. The 5-HT1B Selectivity of CP-93,129, and its lack of 5-HT reuptake blocking properties, suggests that the compound compares favourably with other purported 5-HT1B receptor agonists.