OBJECTIVE There is conflicting evidence as to whether the effect of mannitol on brain bulk arises from haemodynamic, rheologic, or osmotic mechanisms. If mannitol alters cerebral haemodynamics by inducing vasoconstriction, this change should be reflected in cerebral blood flow velocity (CBFV) in the middle cerebral artery (MCA). The purpose of this study was to evaluate the effect of mannitol on CBFV in children. METHODS Children scheduled for intracranial surgery were enrolled. After a loading dose of 10 micrograms.kg-1 of fentanyl, general anaesthesia was maintained with fentanyl (3 micrograms.kg-1.hr-1), 66% nitrous oxide, and isoflurane (0.2-0.5% inspired). Mean and systolic CBFV (Vm and Vs) and pulsatility index (PI) were recorded with a transcranial Doppler (TCD) directed at the M1 segment of the MCA. Mannitol was administered, 1 gm.kg-1 iv over 15 min. The osmolality (Osm), haematocrit (Hct), mean arterial pressure (MAP), heart rate (HR), and TCD variables were recorded before and 15, 30, 45, and 60 min after the mannitol infusion. RESULTS Mannitol infusion resulted in an increase in Osm and decrease in Hct (P < 0.05). Heart rate, MAP and arterial carbon dioxide tensions did not change (P > 0.05) during the measuring period. The Vm did not vary from baseline. The Vs and PI both increased briefly (P < 0.01 at 15 min and P < 0.05 at 30 min) after the mannitol, suggesting an increase in resistance distal to the MCA. CONCLUSIONS The time course of CBFV changes produced by mannitol corresponds with previous animal data concerning cerebrovascular tone. Our results suggest that mannitol briefly increases cerebrovascular resistance and thereby diminishes cerebral blood volume.