OBJECTIVE To compare cardiac output measured in the pulmonary artery and a carotid artery by use of thermal and electrical impedance dilution. Animals-7 fit, clinically normal Standardbreds between 2 and 5 years of age. METHODS Transient changes in electrical impedance and temperature of blood were induced by bolus injections of ice-cold saline hypertonic (6% and 9% NaCl) solutions. Cardiac output was calculated by applying Stewart-Hamilton principles to the indicator dilution transients. Measurements were made during sequential exercise episodes on a level treadmill over approximately an 8-fold range of cardiac output values. RESULTS We detected agreement between cardiac output determined by use of electrical impedance dilution at the pulmonary artery and carotid artery. Cardiac output from thermal dilution measured at the carotid artery exceeded that measured at the pulmonary artery. Cardiac output from the thermal dilution technique exceeded cardiac output from the electrical impedance dilution technique at both locations. CONCLUSIONS The electrical impedance indicator is conserved on first transit; therefore cardiac output measured by electrical impedance dilution at the carotid artery is reliable over a large range of values. Thermal dilution provides a larger estimate of cardiac output, compared with the electrical impedance dilution technique, probably because of a loss of indicator. The transpulmonary electrical impedance dilution technique may have potential for clinical application, particularly in animals in which catheterization of the pulmonary artery is not appropriate or blood loss must be minimized.