In nine patients, with preoperative ICP monitoring, anaesthesia was induced with thiopentone 5 mg kg-1 given over 1 min, followed by pancuronium 0.1 mg kg-1. After manual hyperventilation with nitrous oxide and oxygen for 3 min they were given thiopentone 2.5 mg kg-1 over 30 s (phase 1); 30 s later laryngoscopy was performed and topical analgesia administered to the larynx. Endotracheal intubation was performed 1 min after spraying the cords (phase 2). The measurements continued for a further 5 min during which the patients were mechanically ventilated (phase 3). ICP and intra-arterial pressure were recorded. Although there was a significant decrease (P less than 0.05) in MAP at the end of the second dose of thiopentone, there were no other significant changes in ICP, MAP or PaCO2 throughout the study. In two patients there were transient decreases in cerebral perfusion pressure to less than 60 mm Hg. Although MAP increased in five of the patients during laryngoscopy and intubation, there was no increase in ICP, showing that the MAP was still within the autoregulatory limits.