Effects of prostaglandin E1 (PGE1) or trimethaphan (TMP) on local cerebral blood flow (LCBF) and carbon dioxide (CO2) reactivity were studied in 26 patient undergoing cerebral aneurysm surgery for subarachnoid hemorrhage (SAH) under isoflurane anesthesia. Measurement of LCBF was made using a thermal gradient blood-flow meter. Hypotension was induced with a continuous i.v. infusion of PGE1 or TMP. The dose of PGE1 or TMP was adjusted to maintain the MAP at about 70 mm Hg, and CO2 reactivity was studied during and after PGE1 or TMP administration. PaCO2 was changed by the controlled hyperventilation. CO2 reactivity was evaluated by the following formula: % delta LCBF/delta PaCO2% mm Hg. Hypotensive drugs were discontinued at the completion of aneurysm clipping. Mean arterial pressure (MAP) decreased immediately after the start of either PGE1 or TMP. Heart rate (HR) did not change significantly in either group. LCBF decrease at 30 min after start of TMP administration (p < 0.05 compared with preinfusion value), whereas LCBF was unchanged during PGE1 administration. Neither PGE1 nor TMP affected CO2 reactivity during cerebral aneurysm surgery, whereas CO2 reactivity showed a close correlation with presurgical neurological status: before, rs = -0.523, p < 0.01; during, rs = -0.794, p < 0.01; after, rs = -0.643, p < 0.01. Comparison of the outcomes after aneurysm surgery demonstrated no significant difference between groups; however, there was a close correlation between presurgical neurological status and outcome in both groups (rs = 0.829, p < 0.01). These results suggest that PGE1 is preferable to TMP in inducing hypotension for cerebral aneurysm surgery because this drug maintains both LCBF and CO2 reactivity.