BACKGROUND Isoflurane depresses the electroencephalographic (EEG) activity and exerts part of its anesthetic effect in the spinal cord. The authors hypothesized that isoflurane would indirectly depress the EEG and subcortical response to noxious stimulation in part by a spinal cord action. METHODS Depth electrodes were inserted into the midbrain reticular formation (MRF) and thalamus of six of seven isoflurane-anesthetized goats, and needle-electrodes were placed into the skull periosteum. In five of seven goats, an MRF microelectrode recorded single-unit activity. The jugular veins and carotid arteries were isolated to permit cranial bypass and differential isoflurane delivery. A noxious mechanical stimulus (1 min) was applied to a forelimb dewclaw at each of two cranial-torso isoflurane combinations: 1.1+/-0.3%-1.2+/-0.3% and 1.1+/-0.3-0.3+/-0.1% (mean +/- SD). RESULTS When cranial-torso isoflurane was 1.1-1.2%, the noxious stimulus did not alter the EEG. When torso isoflurane was decreased to 0.3%, the noxious stimulus activated the MRF, thalamic, and bifrontal-hemispheric regions (decreased high-amplitude, low-frequency power). For all channels combined, total (-33+/-15%), delta(-51+/-22%), theta (-33+/-19%), and alpha (-26+/-16%) power decreased after the noxious stimulus (P<0.05); beta power was unchanged. The MRF unit responses to the noxious stimulus were significantly higher when the spinal cord isoflurane concentration was 0.3% (1,286+/-1,317 impulses/min) as compared with 1.2% (489+/-437 impulses/min, P<0.05). CONCLUSIONS Isoflurane blunted the EEG and MRF-thalamic response to noxious stimulation in part via an action in the spinal cord.