Preclinical ultrasound scanners are used to measure blood flow in small animals, but the potential errors in blood velocity measurements have not been quantified. This investigation rectifies this omission through the design and use of phantoms and evaluation of measurement errors for a preclinical ultrasound system (Vevo 770, Visualsonics, Toronto, ON, Canada). A ray model of geometric spectral broadening was used to predict velocity errors. A small-scale rotating phantom, made from tissue-mimicking material, was developed. True and Doppler-measured maximum velocities of the moving targets were compared over a range of angles from 10 degrees to 80 degrees. Results indicate that the maximum velocity was overestimated by up to 158% by spectral Doppler. There was good agreement (<10%) between theoretical velocity errors and measured errors for beam-target angles of 50 degrees-80 degrees. However, for angles of 10 degrees-40 degrees, the agreement was not as good (>50%). The phantom is capable of validating the performance of blood velocity measurement in preclinical ultrasound. (E-mail: email@example.com) (C) 2013 World Federation for Ultrasound in Medicine & Biology.