The effect of nitrous oxide on the activity of the somatosensory system was studied in cats with brain electrodes implanted chronically. The electrodes were implanted in the primary somatosensory cortex, cortical somatosensory radiation, medial lemniscus and midbrain reticular formation. Alterations in the excitability of the primary sensory pathway were assessed by the changes of the input to and output from these brain areas: the response in the medial lemniscus to the stimulation of the skin represented the input to the thalamic relay nucleus and the response recorded in the cortical sensory radiation represented the output from the thalamic relay nucleus. The concentrations of nitrous oxide studied were 50% and 75% in oxygen, and the drug effect was concentration-related. The cortical response to peripheral stimulation was suppressed in amplitude by more than 40% of the control with 75% nitrous oxide, and the response in the cortical radiation was suppressed by 20% of the control with the same dose of nitrous oxide. The response in the cortical radiation to stimulation of the medial lemniscus was suppressed by 20% of the control and the postsynaptic component of the cortical response to the stimulation of the medial lemniscus was suppressed by more than 50% of the control. The multi-unit activity of the brainstem reticular formation was enhanced by nitrous oxide in a dose related manner. The excitability of the thalamic relay nucleus and the primary somatosensory cortex was suppressed by natural slow wave sleep when the reticular multi-unit activity was suppressed, and they were enhanced by paradoxical phase of sleep when the reticular multi-unit activity was enhanced. These findings indicated that the degree of suppression of excitability by nitrous oxide is similar in both the thalamic relay nucleus and sensory cortex, and its action on the brain stem reticular formation is different from that on the primary sensory system. The suppression of sensory functions shown in the present study provides a certain clue to the understanding of the neural basis that though nitrous oxide does not produce deep surgical anesthesia, it does induce potent analgesia and sedation during surgery.