Rat hippocampal synaptosomes preloaded with [3H]serotonin and maintained in a superfusion apparatus were exposed for 3 min to d-fenfluramine or fluoxetine. Both drugs evoked a tritium overflow which was reserpine-sensitive requiring the presence of intact synaptic vesicles. However the two drugs displayed different characteristics: 1) the overflow was immediate with d-fenfluramine whereas the releasing activity of fluoxetine showed a delay of about 2 min; 2) d-fenfluramine-induced overflow was already apparent at 0.15 mumol/l whereas the minimal effective concentration of fluoxetine was 2.5 mumol/l. Their concentration-effect curves were differently shaped, the effect of d-fenfluramine being saturable at 5-20 mumol/l (EC50 about 1 mumol/l) while no saturation was observed with fluoxetine up to 10 mumol/l; 3) only 19% of the tritium overflow evoked by fluoxetine (2.5-10 mumol/l) consisted of true [3H]serotonin, compared with 70% when 0.5 mumol/l d-fenfluramine was used; 4) the releasing action of 0.5 mumol/l d-fenfluramine was completely Ca(++)-dependent, while at higher d-fenfluramine concentrations the Ca(++)-independent overflow became more important. The fluoxetine induced overflow was mainly (70%) Ca(++)-independent; 5) the releasing activity of d-fenfluramine was mainly (80%) blocked by the serotonin uptake blockers indalpine, midalcipram and also fluoxetine whereas fluoxetine-induced overflow was insensitive to inhibition of the serotonin carrier. In conclusion, the releasing activity of d-fenfluramine is already present at a very low concentration (0.5 mumol/l) and at this concentration its mechanism of action was Ca(++)-dependent, together with the requirement of a functional serotonin carrier.(ABSTRACT TRUNCATED AT 250 WORDS)