Background: In this study, we tested the hypothesis that escalating drug concentrations of isoflurane are associated with a significant decline in cerebral blood flow (CBF) in regions sub-serving conscious brain activity, including specifically the thalamus. Methods: Nine human volunteers received three escalating drug concentrations: 0.2, 0.4 and 1.0 MAC end-tidal inhalation. During waking, baseline and the three levels of sedation, a H2 15O PET scan was performed. Results: Isoflurane decreased the bispectral index (BIS) values dose-dependently. Cardiovascular and respiratory parameters were maintained constant over time. No significant change in global CBF was observed. Throughout all three MAC levels of sedation, isoflurane caused an increased regional cerebral blood flow (rCBF) in the anterior cingulate and decreased rCBF in the cerebellum. Initially, isoflurane (0 vs. 0.2 MAC) significantly increased relative rCBF in the medial frontal gyrus and in the nucleus accumbens. At the next level (0.2 vs. 0.4 MAC), relative rCBF was significantly increased in the caudate nucleus and decreased in the lingual gyrus and cuneus. At the last level (0.4 vs. 1 MAC), relative rCBF was significantly increased in the insula and decreased in the thalamus, the cuneus and lingual gyrus. Compared with flow distribution in awake volunteers, 1 MAC of isoflurane significantly raised relative activity in the anterior cingulate and insula regions. In contrast, a significant relative flow reduction was identified in the thalamus, the cerebellum and lingual gyrus. Conclusions: Isoflurane, like sevoflurane, induced characteristic flow redistribution at doses of 0.2-1.0 MAC. At 1 MAC of isoflurane, rCBF decreased in the thalamus. Specific areas affected by both isoflurane and sevoflurane included the anterior cingulate, insula regions, cerebellum, lingual gyrus and thalamus.