Fast cyclic voltammetry (FCV) at carbon fibre microelectrodes was used to monitor endogenous noradrenaline (NA) efflux in superfused slices of bed nucleus of stria terminalis pars ventralis (BSTV) in 'real time'. NA efflux was evoked by local electrical stimulation at bipolar tungsten stimulating electrodes. Confirmation of the identity of the released species as NA was made on the basis of anatomical, electrochemical and pharmacological proofs. Firstly, the signal matched the NA innervation density; efflux of monoamine was greater in BSTV than in the pars dorsalis of the nucleus. Secondly, the voltammogram of the released species was indistinguishable from those of the catecholamines NA and dopamine (DA) but dissimilar to that of the indoleamine serotonin (5-hydroxytryptamine, 5-HT). Thirdly, amine efflux was influenced in a predictable fashion by the drugs tested. Tetrodotoxin (10(-6) M) or omission of Ca2+ from the superfusate reversibly reduced amine efflux by 90.2 and 88.0% respectively. Ro 4-1284 (10(-6) M) decreased amine efflux by 75.8%. Desipramine (5 x 10(-8) M), the selective NA uptake blocker, significantly increased amine efflux and uptake half-life (to 214.3 and 389.5% of control respectively). Fluvoxamine (5 x 10(-7) M) and GBR 12909 (3 x 10(-7) M), blockers of 5-HT and DA uptake respectively, had no effect on amine efflux, although fluvoxamine caused a modest (91.0%) increase in the uptake half-life. Pargyline (2 x 10(-6) M) affected neither efflux nor uptake. The combined anatomical, electrochemical and pharmacological data confirm that the monoamine detected in BSTV by local electrical stimulation was NA. Stimulated NA efflux was stable and reproducible over at least 2.5 h (longest period tested). This study demonstrates the ability of FCV to selectively monitor endogenous NA efflux and uptake in 'real time' and with high spatial resolution.