1. Rat facial motoneurones axotomized on the day after birth were examined morphologically, and their electrical properties were characterized using the whole-cell recording technique in thin slices of the brainstem. 2. About 40% of facial motoneurones were lost within 4 days of axotomy, and only about 20% of the neurones survived 9 days after axotomy. 3. The surviving facial motoneurones examined 4 or 6 days after axotomy were reduced in size, and this was associated with a decrease in their input capacitance. 4. Both the resting potential and the amplitude of action potentials remained unchanged in axotomized facial motoneurones. 5. Facial motoneurones examined 4 or 6 days after axotomy showed an increase in the spike duration. When the preparation was superfused with a Ca(2+)-free solution, the spike duration of axotomized facial motoneurones was shortened, whereas the spike duration of control facial motoneurones was prolonged. 6. The voltage-gated transient K+ current (IA) density was significantly reduced in axotomized motoneurones, whereas the Ca(2+)-dependent transient K+ current (IK, Ca) density was not affected. 7. Voltage-gated Ca2+ currents in facial motoneurones showed inactivation, displaying an initial transient phase followed by a sustained phase. The rate of inactivation of Ca2+ currents was significantly faster in axotomized neurones than in control neurones. 8. A small subpopulation of facial motoneurones examined 4 or 6 days after axotomy had a disproportionately high input resistance and a significantly longer after-hyperpolarization. The probability of this occurring was correlated with the time course of cell death induced by axotomy. 9. It is concluded that facial motoneurones axotomized in neonatal rats comprise two subpopulations. The subpopulation characterized by a markedly high input resistance is suggested to represent the neurones which are at the 'prelethal' stage or in the process of cell death.