The rat rostral raphé nuclei receive catecholaminergic innervation from the locus coeruleus and other areas. In the present study, we investigated noradrenergic modulation of 5-HT release in rat dorsal and median raphé nuclei (DRN and MRN) slices (350 microm thick) superfused with artificial cerebrospinal fluid (aCSF). The raphé was locally stimulated (0.1 ms pulses, 10 mA) and 5-HT release was monitored at carbon fibre microelectrodes using fast cyclic voltammetry. The selective noradrenaline reuptake inhibitor desipramine (50 nM) did not increase stimulated (20 pulses, 100 Hz) 5-HT release but significantly slowed 5-HT reuptake in both DRN and MRN. On short stimulus trains (10 pulses, 200 Hz), the alpha(2)-selective agonist dexmedetomidine (10nM) decreased evoked 5-HT release in DRN and MRN (to 44+/-3 and 43+/-7% of pre-drug values, respectively, at minimum). In both nuclei, this response was antagonised by the selective alpha(2A)-antagonist BRL 44408 (1 microM: P<0.001 vs. dexmedetomidine) but not by the selective alpha(2B/C)-adrenoceptor antagonist ARC 239 (500 nM), the selective 5-HT(1A) antagonist WAY 100635 (100 nM) or the alpha(1)-selective antagonist prazosin (1 microM), suggesting that the effect of dexmedetomidine is wholly attributable to alpha(2A)-receptor activation. The alpha(1)-adrenoceptor agonist phenylephrine (5 microM) significantly decreased 5-HT release (to 49+/-7 and 41+/-4% of pre-drug values in DRN and MRN, respectively). The response was blocked by prazosin (P<0.001) and BRL 44408 (P<0.01) in DRN and by prazosin, BRL 44408 and WAY 100635 (all P<0.05) in MRN, suggesting that the effect of phenylephrine is, under these conditions, only partly mediated via alpha(1)-adrenoceptors. On long stimuli (30 pulses, 10 Hz), BRL 44408 (1 microM) increased evoked 5-HT efflux to 187+/-17 and 178+/-2% of pre-drug values in DRN and MRN, respectively (both P<0.001 vs. vehicle). Collectively, these data show that activation of both alpha(1) and alpha(2A)-adrenoceptors can decrease stimulated 5-HT release in the rostral raphé nuclei. Since the effect of dexmedetomidine was not antagonised by prazosin, we suggest that its effect was mediated directly, possibly through alpha(2A) receptors located on 5-HT cell elements, and not transduced indirectly through alpha(1)-adrenoceptor activation, as previously suggested by others.