This study reports the effect of chronic paroxetine (10 mg/kg p.o., 21 days) on 5-HT1B and 5-HT1D autoreceptors controlling stimulated 5-HT efflux in slices of rat dorsal raphe nucleus. Electrically evoked 5-HT (10 pulses, 200 Hz, 0.1 ms, 10 mA) was measured using fast cyclic voltammetry. 5-HT efflux was inhibited by CP 93129 (10 nM-10 mu M) and by sumatriptan (1 nM-1 mu M) agonists at 5-HT1B and 5-HT1D receptors, respectively. Chronic paroxetine did not, initially, appear to alter the sensitivity of the 5-HT1B autoreceptors to CP 93129. However, when constructed in the presence of WAY 100635 (10 nM) the selective and silent 5-HT1A antagonist, there was a significant (P < 0.001) rightward shift of the CP 93129 concentration-response curve in the paroxetine-treated rats but not in the controls, implying a desensitisation of the 5-HT1B autoreceptor by paroxetine. Chronic paroxetine did not affect the sumatriptan concentration-response curve, even with WAY 100635 present, implying that there was no (de)sensitisation of the 5-HT1D autoreceptor. These data suggest that chronic paroxetine treatment may desensitise 5-HT1B autoreceptors in the dorsal raphe nucleus but that this effect is unmasked only when the dominant 5-HT1A autoreceptor control is antagonised. (C) 2000 Elsevier Science Ltd. All rights reserved.