The cerebrovascular response to CO2 has been reported to be preserved during propofol anesthesia, but no comparison with awake control values has been made, and the additional influence of N2O has not been investigated. Using the noninvasive technique of transcranial Doppler ultrasonography, this study investigated the cerebrovascular response to varying levels of Pa(CO2) while awake and during anesthesia with propofol and propofol/N2O. Seven adults without systemic diseases undergoing nonneurologic surgery were studied. A pulsed-wave Doppler monitor was used to measure the mean middle cerebral artery flow velocity (V(mca)) during varying levels of Pa(CO2) (25-55 mmHg) under the following conditions: 1) awake; 2) propofol 2.5 mg.kg-1 bolus followed by continuous infusion of 150 mug.kg-1.min-1; and 3) propofol as in the condition above plus 70% N2O. During the awake study condition, hypocapnia was induced by voluntary hyperventilation, and hypercapnia was induced with rebreathing of 7% CO2 in a closed circuit. During the anesthetized study conditions, hypocapnia and hypercapnia were induced by adjustment of minute ventilation. A minimum of five to six simultaneous V(mca) and Pa(CO2) measurements were obtained under each of the study conditions. Systemic blood pressure was monitored via a radial arterial catheter, and phenylephrine was administered if mean arterial blood pressure decreased below 60 mmHg (phenylephrine was used in three of five patients in the propofol-N2O group). Linear regression and analysis of covariance were used for statistical analysis of V(mca)-Pa(CO2) relationships. At all levels of Pa(CO2), the V(mca) recorded during propofol alone and propofol-N2O was significantly less than that recorded during the awake state (the reduction ranged from 36% during hypercapnia to 45% during hypocapnia). Compared to the CO2-reactivity slope during the awake state of 3.2 +/- 0.2% mmHg-1 (mean +/- SE, n = 7), the slopes during propofol (n = 7) and propofol-N2O (n = 5) were 2.1 +/- 0.2% and 2.5 +/- 0.2% respectively (P < 0.05 compared to awake). We conclude that propofol has vasoconstrictive properties on the cerebral vasculature and that the addition of N2O does not influence the vasoconstriction or the CO2 reactivity.