The authors evaluated a thermodilution catheter designed to continuously measure cardiac output (CO). A 10 cm long surface heating element is positioned in a Swan-Ganz catheter corresponding to a right atrial-ventricular site. Heat is repetitively deposited into flowing blood in a unique, pseudorandom binary form. Small temperature fluctuations are sensed with a high performance thermistor and correlated with the heat input pattern, from which CO is determined. Seven adult sheep were anesthetized and instrumented for both continuous and standard cold bolus injection thermodilution (COM1) flow measurements. Heart rate and blood volume were adjusted to vary CO from 1.5 to 13.2 L/min. Continuous measurements correlated well with triplicate COM1 determinations (Sy,x = 0.56, r = 0.967) that improved with experience (Sy,x = 0.38, r = 0.99 for the last three animals). The surface heat transfer coefficient was measured in water (catheter parallel to flow). Results agreed well with a standard cylinder-in-crossflow correlation. The right ventricle heating element surface temperature was predicted for several CO and heating combinations. Worst case results yielded a 5.8 degrees C surface temperature elevation, suggesting that thermally induced damage is unlikely. Results suggest this catheter provides accuracy at least comparable to that of standard cold bolus injection methods, with no heat induced damage to blood.