So far, the mechanisms underlying the action of selective serotonin reuptake inhibitors, such as fluoxetine, are not completely understood. Thus, to clarify if fluoxetine has any effect on noradrenergic transmission, we measured the spontaneous firing rate of noradrenergic neurons in the locus coeruleus both in vivo and in vitro using single-unit extracellular recordings. In anesthetized rats, fluoxetine (2.5-20 mg/kg, i.v.) reduced the firing rate in a dose-dependent manner, reaching a maximal inhibition of 55 +/- 5% with respect to the basal value. This effect was not only completely reversed by the alpha(2)-adrenoceptor antagonist, RX 821002 (0.2 mg/kg, i.v.), but also prevented by previous application of both idazoxan (0.05 and 0.1 mg/kg, i.v.) and RX 821002 (6.25 microg/kg, i.v). Furthermore, when noradrenaline was depleted from axon terminals by means of the injection of alpha-methyl-DL-tyrosine (250 mg/kg, i.p.) 24 h prior to the experiment, fluoxetine failed to inhibit locus coeruleus activity. In rat brain slices, perfusion with fluoxetine (100 microM for 5 min) did not modify the firing rate of locus coeruleus neurons (n = 7). We conclude that fluoxetine inhibits locus coeruleus neurons in vivo through a mechanism involving noradrenaline interacting with alpha(2)-adrenoceptors. However, the lack of effect on brain slices would seem to indicate that afferents to the nucleus may be involved in the observed inhibitory effect.