A comparison of two volumetric blood flow measurement techniques, CVI-Q (based on time domain correlation) and ultrasonic flowmetry, has been performed in vitro and in vivo. A pulsatile flowpump was used to simulate carotid and femoral type waveforms which were measured simultaneously using the two methods. Five dogs had their common carotid and femoral arteries exposed, and the instantaneous maximum volume flow and the mean flow were measured. Each vessel was partially occluded halfway through the experiment, simulating a 90% stenosis. In vitro, both techniques achieved absolute errors below +/- 5% for flow rates over 100 mL/min, but ultrasonic flowmetry had statistically significant larger errors for slower flow rates. In vivo correlation coefficients ranging from 0.73 to 0.95 were obtained with regression line slopes close to unity. The two techniques were in reasonable agreement, but with standard deviations of 20% to 28%. These studies indicate that noninvasive CVI-Q measurements of blood flow in the carotid and femoral arteries are linear and accurate compared to invasive ultrasonic flowmetry.