We used intracerebral microdialysis to study the role of raphe and presynaptic serotonin (5-HT) autoreceptors in the effect of the selective 5-HT reuptake inhibitor, paroxetine, on 5-HT release from ventral hippocampus of anaesthetised rats. In addition, we have tested the ability of pindolol, a non-selective beta-adrenergic/5-HT(1A) receptor antagonist, to alter the response of hippocampal 5-HT to paroxetine. Doses of paroxetine with maximal effects were near to three-fold less effective when administered systemically than after local infusion at increasing extracellular 5-HT in ventral hippocampus. Moreover, systemic paroxetine treatment resulted in a marked decrease of the extracellular 5-HT in the hippocampus when 5-HT reuptake was blocked with paroxetine 3 microM applied locally, thereby evidencing that systemic treatment induced a decrease of 5-HT release in the neuronal terminal. A similar drop was observed when paroxetine 3 microM was perfused into the median raphe, a region that contains the cell bodies of the neurons innervating the ventral hippocampus. Racemic (+/-)-pindolol (10 mg/kg, s.c.) completely blocked the paroxetine-induced decrease in 5-HT release from rat hippocampus. In addition, the infusion into median raphe of (-)-pindolol, the isoform with highest antagonist activity, at concentrations of 10 microM and 100 microM was able to partially block the decrease of hippocampal 5-HT release after systemic paroxetine. However, perfusion of (-)-pindolol into the hippocampus was without effect on local 5-HT release. These data suggest that pindolol acts preferentially through the blockade of somatodendritic 5-HT(1A) autoreceptors to restore the decline in 5-HT outflow in rat forebrain following systemic administration of selective 5-HT reuptake inhibitors.