The peak blood velocity is used in important diagnostic applications, e.g. for determining the stenosis degree. The peak velocity is typically assessed by detecting the highest frequency in the Doppler spectrum. The selected frequency is then converted to velocity by the Doppler equation. This procedure contains multiple potential sources of error: the peak frequency selection is sensitive to noise and affected by spectral broadening, and the frequency to velocity conversion is altered by the Doppler angle uncertainty. The result is an inaccurate estimate. In this work we propose a new method that removes the aforementioned errors. By exploiting a mathematical model of the Doppler spectrum the exact frequency to be converted to velocity, with no need of broadening compensation, is determined. The angle ambiguity is solved by calculating the Doppler spectra backscattered from two different receive apertures. The proposed methods uses, in transmission and receive, defocused steered waves that produce a wide sample volume. This includes the whole vessel section making the probe positioning quick and easy. The method, validated through Field II simulations and phantom experiments, featured a mean error lower than 1%.