Two noninvasive methods of cardiac index (Q) determination, pulsed Doppler echocardiography (QDop), and CO2 rebreathing (QCO2) were compared to dye dilution in eight normal volunteers. Measurements of Q were made by dye dilution (QDD) and the two noninvasive techniques under the following conditions: supine rest, after inflation of cuffs around the thighs, 35-degree head-up tilt, supine rest repeated, during constant isoproterenol infusion, and after an intravenous bolus of propranolol. When mean Q values of the group for each intervention were compared, close agreement was observed between dye dilution and each noninvasive method (QDOP = 1.08 QDD - 0.07 L/min X m2, r = 0.99, SEE = 0.02; QCO2 = 0.68 QDD + 0.84 L/min X m2, r = 0.97, SEE = 0.02). Fair correlations were achieved when all 48 paired observations were analyzed (QDop = 1.00 QDD + 0.17 L/min X m2, r = 0.89, SEE = 0.17; QCO2 = 0.71 QDD + 0.77 L/min X m2, r = 0.79, SEE = 0.20). However, the Doppler ultrasound technique was significantly more precise in quantitating changes in Q in individual patients. These results demonstrate that estimates of Q can be made with reasonable accuracy by either CO2 rebreathing or Doppler echocardiography. However, the Doppler technique is a much more accurate means of quantitating acute changes in Q.