SUMMARY The purpose of this study was to compare the thermodilution technique for estimation of cardiac output with the indocyanine green dye dilution technique at flows between 10 and 39 L/min in halothane-anesthetized horses. The estimation of area of dye dilution cardiac output curves was made by using the fore-’n-aft (fa) triangle method. This shorthand technique was compared with logarithmic exponential extrapolation and summation (extrapolated area), using 64 cardiac output curves. Then, 256 simultaneous thermodilution measurements were compared with dye dilution measurements calculated by use of the fa technique. Forty milliliters of iced 0.9% NaCl solution containing 15 mg of indocyanine green dye was used as the indicator. This was delivered in < 1 second to the right atrium, using a power injector. A thermistor positioned in the pulmonary artery detected the thermal indicator. Blood was withdrawn from the carotid artery through a densitometer cuvette to measure the dye concentration. The fa estimations of area were higher than those determined by use of extrapolated area. A multiplicative adjustment of 0.837 was estimated. On average, thermodilution estimates of cardiac output exceeded the adjusted fa determinations. Using a weighted linear regression, we determined the following calibration adjustment: thermal dilution cardiac output/1.048 = indocyanine green dye dilution cardiac output.