1. Intracellular recordings were made from antidromically identified motoneurones in transverse spinal cord slices from neonatal (12-16 day) rats. 2. Superfusion of (+/-)-baclofen (0.5-50 microM) reduced the excitatory postsynaptic potentials (e.p.s.ps) and inhibitory postsynaptic potentials (i.p.s.ps) evoked by dorsal root or dorsal root entry zone stimulation in a concentration-dependent manner; the calculated EC50 was 2.4 microM. Baclofen in comparable concentrations also reversibly eliminated spontaneously occurring e.p.s.ps and i.p.s.ps. 3. (-)-Baclofen was more effective as compared to baclofen in reducing the synaptic responses, whereas (+)-baclofen at concentrations as high as 50 microM was ineffective. 4. Baclofen (less than 5 microM) attenuated the synaptic responses without causing a significant change of passive membrane properties and depolarizations induced by exogenously applied glutamate. In addition to synaptic depression, baclofen (greater than 5 microM) caused a hyperpolarization associated with decreased membrane resistance in some of the motoneurones; the glutamate responses were also attenuated. 5. Baclofen reversibly depressed the spike after-hyperpolarization of the motoneurones. 6. GABA (1-10 mM) depressed synaptic transmission and depolarized or hyperpolarized motoneurones. While potentiated by the uptake inhibitor nipecotic acid, the synaptic depressant effect of GABA was not antagonized by bicuculline. 7. The synaptic depressant effect of baclofen was neither blocked by GABAA antagonists bicuculline and picrotoxin (10-50 microM) nor by the GABAB antagonist phaclofen (0.1-1 mM). 8. It is suggested that baclofen depresses excitatory and inhibitory transmission in rat motoneurones by primarily a presynaptic mechanism in reducing the liberation of chemical transmitters from nerve endings via a phaclofen-insensitive GABAB receptor.