1. Cats were prepared for chronic recording from neurons in pars oralis and pars interpolaris of the trigeminal spinal nucleus. Electrodes were implanted into canine teeth for electrical stimulation and the digastric muscle for recording electromyograms. 2. Recordings were made from the animals when they were awake and unrestrained as well as when they were lightly anesthetized. Some neurons were studied under both conditions. 3. In an awake animal, single tooth-pulp stimuli of 0.1 ms duration and < or = 1 mA intensity produced no aversive behavior. 4. The response of trigeminal brain stem neurons in the awake animal to such stimuli consisted of short (approximately 3 ms)- and long (approximately 25 ms)-latency discharges whose thresholds suggested that they were both due to inputs from fast conducting primary afferent fibers. 5. Light anesthesia reduced the number of impulses in both components and in most cases completely abolished the long-latency component evoked by low-intensity stimuli. The threshold of the short-latency component was little affected by light anesthesia. It is postulated that the short-latency component is mediated by a monosynaptic input from primary afferent fibers and the long-latency component by a polysynaptic input from these same fibers. 6. All neurons that responded to tooth-pulp stimulation had inputs from other orofacial sites both in the awake and lightly anesthetized states. After light anesthesia, these receptive fields were altered in only 3 out of 15 neurons. 7. The majority of neurons (18 out of 20) were not spontaneously active in the awake animal. Spontaneous activity in the other two was reduced by light anesthesia. 8. The threshold of the digastric reflex evoked by tooth-pulp stimulation was not altered by light anesthesia, but the size of the response was reduced. 9. The effects of changing the level of anesthesia from deep to light (i.e., without and with reflex withdrawal to squeezing a paw) on the responses to tooth-pulp stimulation were also studied. Decreasing the anesthetic depth tended to decrease the thresholds and increase the magnitude of both the short- and long-latency neuronal responses and the short-latency digastric response.