A first plurality of MR signals from a patient's tissue at respectively corresponding successive first time increments extending over a first time interval including a substantial majority of a subject's cardiac cycle is acquired and analyzed to define a second time interval, shorter than the first time interval, during the cardiac cycle whereat there is a relatively steep rise in signal magnitudes as a function of time (e.g., corresponding with systole and diastole events of the cardiac cycle). A second plurality of MR signals is then acquired from tissue of the patient at respectively corresponding successive second time increments during the second time interval, the second time increments being substantially shorter than said first time increments. Image data representing at least one contrast-free image of flowing fluid vessels is generated based on the second plurality of MR signals.