The nitrous oxide method for determining cerebral blood flow as originally described by Kety and Schmidt (1) has been unsuitable for use in children because of the fairly large amounts of blood withdrawn in its performance and because of its requirement of a greater degree of active cooperation than can be expected from most children under the circumstances of the procedure. Although the integrated sample modification of Scheinberg and Stead (2) permits a considerable reduction in the amount of blood drawn, it loses a valuable check on potential errors arising from defects in the respiratory system or the presence of extracerebral contamination of the internal jugular venous blood. Furthermore, timing and dilution problems associated with the Scheinberg modification lead to errors unless extensive precautions are taken (3). Recently, Baird and Garfunkel (4) have employed a modification of the original method in studies of the cerebral blood flow in infants and children. Blood requirements were reduced by the use of micro-gas analytical techniques (5, 6), but the necessary cooperation was achieved only by sedation with barbiturates. Although these drugs, in semi-narcotic doses, have been shown to be without effect on the cerebral circulation and oxygen consumption in adults (7), this has not been established in children. When anesthetic doses are employed, a marked reduction in cerebral oxygen consumption has been observed