1. Experiments were performed on the brain stem-spinal cord preparation of newborn rats, in which the phrenic and hypoglossal nerves continue to show rhythmic respiratory activity in vitro, in order to compare the effects of serotonin (5-HT) on both activities and to analyse the mechanisms responsible for the depression by 5-HT of the hypoglossal activity. 2. Under control conditions, simultaneous recordings of the inspiratory discharges of hypoglossal and cervical roots showed that the two bursts did not start simultaneously and had different patterns (time-to-peak and peak values); this suggests that both pools of motoneurons did not share the same central drive(s). 3. Adding 5-HT and related agents to the bathing medium delayed and depressed the hypoglossal inspiratory discharge via activation of 5-HT2 receptors since these effects were elicited by 5-HT2 agonists (alpha-methyl-5-HT and 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane-HCl (DOI)) but not by 5-HT1 agonists (RU 24969 and (+/-)-8-hydroxy-2-(di-N-propylamino)tetralin hydrobromide (8-OH-DPAT)). The 5-HT depression of the hypoglossal discharge was prevented by applying a pretreatment with a specific 5-HT2 antagonist (ketanserin). Parallel to the hypoglossal discharge decrease, 5-HT elicited a permanent cervical root discharge along with a persistent inspiratory bursting. Adding the 5-HT precursor L-tryptophan to the bathing medium depressed the hypoglossal (XII) discharge without affecting the cervical one. 4. Local application of 5-HT within the hypoglossal motor nucleus decreased the hypoglossal output, revealing that the 5-HT depression of the hypoglossal discharge was at least partly mediated by the 5-HT effects at the level of the motoneurons. Local application of 5-HT within the cervical motor nucleus elicited a permanent firing in the cervical root with a persistent inspiratory bursting. 5. Intracellular analysis confirmed the existence of differences in central respiratory drive between cervical and hypoglossal motoneurons under control conditions, as well as differences in response to 5-HT. All the hypoglossal motoneurons became silent under 5-HT bathing, and showed no change in the input membrane resistance, a moderate depolarization, and a delayed central respiratory drive with a decreased amplitude. The cervical motoneurons became more active during inspiration, despite a decrease in the amplitude of the central respiratory drive, which was compensated for by a large depolarization and an increased input membrane resistance. Some cervical motoneurons even fired at a low rate during expiration.(ABSTRACT TRUNCATED AT 400 WORDS)