1. To compare the responsiveness of lumbar spinal neurons to peripheral sensory stimuli under normal physiological conditions and under halothane anesthesia, we performed a study in sheep that were prepared chronically. This permitted recordings to be made in the same animals either when they were awake and free from recent surgery, drugs, and training and only partially restrained or when they were anesthetized with halothane. 2. We recorded 261 units in dorsal and ventral horns under conscious conditions. Of these, 19% had no detectable receptive field (RF) and 44% had responses dominated by proprioceptive inputs; these units were not investigated in detail. The remaining 96 neurons (37%) had clearly defined cutaneous RFs. Of these, most (72%) had wide-dynamic-range (WDR; convergent, multireceptive) properties, 19% were low-threshold mechanoreceptive (LTMR), and 9% were high-threshold mechanoreceptive (HTMR). These units with cutaneous RFs were investigated in greater detail. 3. The spontaneous activity under these awake conditions was low (< 4 spikes/s) for nearly all units in all three categories. The mechanical threshold of the most sensitive (central) part of the cutaneous RF was assessed with von Frey bristles. Thresholds were < 5 mN for all LTMR neurons, < 1-30 mN for WDR neurons, and > 80 mN for HTMR neurons. The size of the low-threshold cutaneous RFs was significantly larger for WDR neurons (mean 46 cm2) and HTMR neurons (45 cm2) than for LTMR neurons (24 cm2). The RFs were distributed all over the ipsilateral hindlimb. Large RFs were mostly proximal, whereas small RFs were distributed relatively evenly over the limb. 4. Recordings were made from a further 165 units while the animals were under halothane anesthesia. With 86 neurons having cutaneous peripheral RFs, the proportions having LTMR, HTMR, or WDR characteristics were very similar to those in awake animals. Under halothane the ongoing activity of WDR units was slightly (but significantly) less. The threshold to von Frey bristle stimulation was significantly higher only for WDR units, in both dorsal and ventral horns. The mean size of cutaneous RFs was significantly larger in all classes of units recorded under halothane anesthesia. For WDR units this was true for cells in both dorsal and ventral horns. This effect on mean values was due to a larger proportion of units with very large fields under anesthesia, particularly in the dorsal horn. 5. Comparison of the data from conscious animals with published results of acute experiments indicates that acute recording conditions do not distort the relative distribution and resting characteristics of these three functional categories of lumbar spinal neurons as much as might have been expected. 6. Halothane does not have major effects on the resting sensory responsiveness of spinal neurons with cutaneous RFs. The increase in RF area, which contrasts with most results from acute studies, is likely to be due to a dampening of descending inhibitory control mechanisms.