This study was carried out to evaluate the effects of perioperative indomethacin on intracranial pressure (ICP), cerebral blood flow (CBF), and cerebral metabolism. Twenty patients subjected to craniotomy for supratentorial cerebral tumors were anesthetized with thiopental, fentanyl, nitrous oxide, and isoflurane. A PaCO2 level averaging 4.8 kPa (median) was achieved. The patients were randomized to intravenous indomethacin 50 mg or placebo administrated after exposure of the dura. ICP was measured continuously subdurally with a 22-gauge canula connected to a transducer. CBF and the arteriovenous difference of oxygen (AVDO2) were measured twice, before and after indomethacin/placebo administration. A significant decrease in ICP from 6.5 to 1.5 mm Hg (median) was found after indomethacin administration. This decrease was caused by a significant decrease in CBF associated with a significant increase in AVDO2. Indomethacin did not affect the cerebral metabolic rate of oxygen, the arteriovenous difference of lactate, or the lactate/oxygen index, suggesting that indomethacin did not provoke global cerebral ischemia. In the indomethacin group, dura was sufficiently relaxed in eight of nine patients and dura was opened without the occurrence of cerebral swelling. In one patient, mannitol treatment was necessary to prevent dural tightness. In the placebo group, mannitol supplemented with hypocapnia was applied in five patients. These findings suggest that perioperative treatment with indomethacin is an excellent treatment of intracranial hypertension during normocapnic isoflurane anesthesia for craniotomy.