We investigated the cerebral haemodynamic effects of 1 MAC desflurane anaesthesia in nine male patients scheduled for elective coronary bypass grafting. For the measurement of cerebral blood flow (CBF) a modified Kety-Schmidt saturation technique with argon as inert tracer gas was used. Measurements of CBF were made before induction of anaesthesia and 30 min after induction under normocapnic, hypocapnic and hypercapnic conditions in sequence. Changes in mean arterial pressure after induction of anaesthesia and during the course of the study were minimized using norepinephrine infusion. In comparison with the awake state under normocapnic conditions, desflurane reduced mean cerebral metabolic rate of oxygen (CMRO2) by 51% and mean cerebral metabolic rate of glucose (CMRglc) by 35%. Concomitantly, CBF was significantly reduced by 22%; jugular venous oxygen saturation (SjvO2) increased from 58 to 74%. Hypo- and hypercapnia caused a 22% decrease and a 178% increase in CBF, respectively. These findings may be interpreted as the result of two opposing mechanisms: cerebral vasoconstriction induced by a reduction of cerebral metabolism and a direct vasodilator effect of desflurane. CBF alterations under variation of PaCO2 indicate that cerebrovascular carbon dioxide reactivity is not impaired by application of 1 MAC desflurane.