The effect of manipulating the activity of central 5-hydroxytryptamine (5-HT) neurones on extracellular 5-HT in ventral hippocampus of the chloral hydrate-anaesthetized rat was studied using the brain perfusion method, microdialysis. Basal levels of 5-HT in the dialysates were close to the detection limits of our assay using HPLC with electrochemical detection. However, addition of the selective 5-HT reuptake inhibitor citalopram (10(-6) M) to the perfusion medium produced readily measurable amounts of dialysate 5-HT. Citalopram, therefore, was used throughout our experiments. Hippocampal dialysate levels of 5-HT sharply declined over the first hour after dialysis probe implantation, but then became constant. This stable output of 5-HT was reduced by 57% in rats treated 14 days previously with intracerebroventricular injections of the 5-HT neurotoxin 5,7-dihydroxytryptamine. Electrical stimulation (1-ms pulse width, 300 microA, 2-20 Hz) of the dorsal raphe nucleus for 20 min caused a rapid rise in hippocampal 5-HT output, which immediately declined on cessation of the stimulus and was frequency-dependent. Addition of tetrodotoxin (10(-6) M) to the perfusion medium reduced 5-HT levels to 75% of predrug values. Injection of the 5-HT1A agonist 8-hydroxy-2-(di-n-propylamino)tetralin (0.5 and 2.5 micrograms) into the dorsal raphe nucleus caused a dose-related fall in hippocampal output of 5-HT compared to saline-injected controls. We conclude from these data that the spontaneous output of endogenous 5-HT into hippocampal dialysates, measured under our experimental conditions, predominantly originates from central 5-HT neurones and changes in accordance with their electrical activity.