A 4 F thermodilution catheter for measuring cardiac output was evaluated for accuracy and linearity in the laboratory and by comparison with the dye dilution method in infants and children following cardiac surgery. When 2 ml of 0°C injectate were used, the correlation of computer determined flows to calibrated pump flows, over a range encountered clinically, wasr= 0.998. The means of triplicate determinations by both the thermal and dye methods were compared in 8 of 25 patients and the comparison found to be favorable (r= 0.976). The complications of thermodilution catheter placement are described and related to the need for post-surgical chest x-ray and thermodilution recordings.The simplicity of the thermodilution technique and other advantages over the dye method in children, such as repeatability, and ease of calibration are discussed in relation to the increased flexibility in management which accrues.