While there has been wide general acceptance of Doppler methods that use the simplified Bernoulli relationship to estimate pressure gradients across stenotic orifices, there is still ongoing controversy related to potential sources of error in the method. In this study we tested accuracy of ultrasound Doppler measurements of flow velocity when compared with the gold standard of laser light Doppler anemometry in a pulsatile flow model of pulmonic stenosis in vitro. We tested two commercially available Doppler systems and examined steered and nonsteered, parallel, and off-axis and anglecorrected velocity determinations using continuous-wave and high-pulse repetition frequency (HPRF) methods. We also examined the potential range of error in the simplified Bernoulli method. One hundred and twenty individual flow states were examined with three stenotic valve orifices (3.0, 1.0, and 0.5 cm2 flow area) to measure velocities up to 620 cm/sec. A very high correlation coefficient was obtained for the comparison of laser Doppler anemometric and ultrasound velocity recordings by the nonsteered continuous-wave technique (r = .99, SEE = 17.9 cm/sec), but there was a tendency for underestimation of higher velocities when the transducer was positioned at 30 degrees and the ultrasound beam was steered so as to be parallel to the visualized flow jet (r = .98, SEE = 29.6 cm/sec). The HPRF ultrasound Doppler technique was also highly accurate in this optimized setting for measuring velocities (r = .99, SEE = 17 cm/sec), but also slightly underestimated the highest velocities. Our results also verified the accuracy of the simplified Bernoulli equation for converting instantaneous velocity measurements to estimated peak instantaneous gradient (r = .97, SEE = 8.4 mm Hg).