The use of phase-encode grouping (PEG) allows acquisition of a complete cardiac cine in a single breath hold, eliminating respiratory artifacts and improving edge definition. One approach to quantitative magnetic resonance (MR) flow measurements in pulmonary, coronary, and renal arteries uses phase velocity mapping with a PEG, single breath-hold, cine technique. Phantom studies were designed to evaluate the accuracy of MR phase velocity mapping (MRV) with PEG in continuous and pulsatile flow measurements in small vessels. Experiments were performed on a 1.5-T MRI system and results were compared with those obtained by timed volume measurements and a transit-time ultrasound flowmeter. Flow velocity, pulsatile flow wave form, pulsing rate, and PEG size were varied. For continuous flow, MRV measured velocities correlated well with those calculated from timed volume measurements (r > 0.998) under mean spatial velocities ranging from 20-100 cm s-1. The temporal mean velocity difference between MRV and ultrasound flowmeter measurements was less than +/- 2.9 cms-1 under pulsatile flow conditions with PEG sizes of 5 or less, however, the mean velocity difference was > 7 cm s-1 for PEG sizes of 8 or more. Accurate MR velocity measurement with PEG in small vessels under conditions of pulsatile flow is possible if an appropriate PEG size is selected.