To estimate axial velocity for vessels that are so narrow that their width is comparable to the minimum length of the range cell, it has been customary to substitute the maximum Doppler frequency fmax into the classical Doppler equation [eqn (1)]. It was shown here that this ignores transit time broadening, which can lead to significant errors at large beam-to-flow angles. We use a relation [eqn (2)], which does take this broadening into account, and give in vivo experimental proof of this allows accurate estimation of the axial velocity even when the range cell extends across the whole vessel lumen. It may be concluded, therefore, that, by using the above procedure for taking transit time spectral broadening into account, more accurate velocity estimates can be obtained for laminar flow than is possible with the current method using the classical Doppler equation uncorrected for broadening.