1. In cats anaesthetized with alpha-chloralose, micro-electrodes have been used to record the discharge patterns of single neurones in the region of the nucleus interpositus. 2. Almost all cells tested could be antidromically invaded following electrical stimulation of the contralateral red nucleus, showing that they were cerebellar efferent neurones. 3. A little over half of the interpositus neurones were spontaneously active, usually at rates of less than 20 impulses/sec. 4. About 40% of the cells had no spontaneous activity, although they gave brisk responses to electrical stimulation of cutaneous nerves. Such silent units were encountered most frequently in the earlier stages of an experiment, but a number were found more than 15 hr after the beginning of an experiment. 5. Stimulation of cutaneous and mixed nerves of the fore and hind limbs provoked impulse discharges of the cells and also produced phases of deceleration of the resting discharge of spontaneously firing cells. 6. The typical response of an interpositus neurone consisted of a short latency (6-35 msec) discharge, usually separated from a long latency (50-500 msec) discharge by a period of inhibition or return to the resting discharge rate. The two phases of excitation appeared to be independently generated, since in a number of cells one phase appeared without the other. In addition, the later phase of excitation was abolished in all cells tested by a small dose of pentobarbitone which produced very little effect on the earlier phase. The long latency response was quantitatively much greater, sometimes consisting of 50 or more impulses in a response which lasted several hundred msec, but was very variable from one trial to another. 7. The long latency discharge and sometimes the preceding inhibition could readily be mimicked by single shock stimulation of the region of the contralateral inferior olive. Short latency discharges were, however, rarely evoked by olivary stimulation. 8. It is suggested that the short latency responses of the interpositus neurones were a result of synaptic excitation via cerebellar afferents, while the ensuing inhibition was a result of post-synaptic inhibition resulting from the Purkinje cell excitation due to the afferent volleys. It is suggested that the long latency excitation is due to the afferent volleys. It is suggested that the long latency excitation is due at least in part to disinhibition resulting from long pauses in Purkinje cell firing following their activation by climbing fibre afferents. 9. The possibility that these long latency responses have a physiological significance in relation to locomotion is discussed.