The authors developed and experimentally verified expressions that describe the frequency response of prospectively gated phase-contrast magnetic resonance velocity measurements. Both interleaved and noninterleaved phase-contrast techniques were evaluated. The primary determinants of the frequency response were (a) the number of interleaved acquisitions (N), (b) the time between acquisitions (delta T), and (c) the degree of balance between the first moments of the velocity-encoding gradients. To quantify the last factor, an imbalance parameter (U) was defined. Depending on the chosen implementation and U, deviations from the ideal frequency responses were predicted and observed. The expressions also revealed an advantage of interleaved acquisitions that use a one-sided gradient configuration: no changes in the frequency response. The effects of concurrently encoding orthogonal velocity components with a Hadamard four-point scheme were examined.