Because it is common to manipulate PaCO2 during neurosurgery, it is essential to characterize the relationship between cerebral blood flow (CBF) and changes in PaCO2. The purpose of this study was to investigate the effects of propofol-N2O anesthesia on the CBF response to changes in PaCO2 in healthy subjects. In seven patients, anesthesia was induced with propofol 2.0-2.5 mg/kg and then maintained with a propofol infusion of 12 mg.kg-1.h-1 for 10 min and then 9 mg.kg-1.h-1 for 10 min and then was reduced to 3-6 mg.kg-1.h-1 for the remainder of the study. The subjects' lungs were ventilated with N2O in O2 (FIO2 0.3) to the end-tidal CO2 present before anesthesia, and then CBF was measured using intravenous 133Xe and ten scintillation counters, five over each cerebral hemisphere. ETCO2 then was increased to 50 mmHg and CBF measurement repeated; ETCO2 then was reduced to 30 mmHg and CBF measurement repeated. Concurrent with each CBF measurement, arterial blood was sampled for PaCO2 and hemoglobin measurement. CBF at normocapnia (PaCO2 42 +/- 2 mmHg) was 33 +/- 7 ml.100 g-1.min-1, which increased to 58 +/- 10 ml.100 g-1.min-1 and decreased to 19 +/- 4 ml.100 g-1.min-1 on increasing PaCO2 (53 +/- 4 mmHg) and decreasing PaCO2 (31 +/- 2 mmHg), respectively. Both the PaCO2 and CBF values were statistically different from those measured at any other time (CBF P less than 0.002, PaCO2 P less than 0.001). The slope of CBF versus PaCO2 was 1.56 ml.100 g-1.min-1.mmHg.(ABSTRACT TRUNCATED AT 250 WORDS)