Left-to-right intracardiac shunts with pulmonary (Qp)to systemic (Qs) flow ratios of >1.5: 1 are considered significant. Accordingly, this study was doneto assess if velocity-encoded, phase-difference cine magnetic resonance imaging (MRI) could reliably determine whether or not patients had Qp/Qs ratios of >1.5: 1. Twenty patients underwent MRI examinations followed within 1to 2 hours by cardiac catheterization. To measure flow, velocity-encoded, phase difference MRI sequences were positioned perpendicula rly across the proximal aorta and pulmonary artery (PA). These scans were 10mm slices with 256×256 matrices, had fields of view ranging from 32 to 45cm (yielding voxel sizes of 1×10×1.3–1.7mm), a flip angle of 40’, a repetition time of 19.5msec, and an echo time of 11msec. Phase encoding grouping sizes of 2 or 3 were used to acquire multiple phase encoding steps for each frame during each cardiac cycle, thereby reducing scan times. All scans were performed with prospective ECG and respiratory gating. Qp and Qs were calculated by multiplying the heart rate by the sum of the flow for all frames of the cardiac cycle in the PA and aorta, respectively. During catheterization, hydrogen appearance times were used to determine the presence of a left-to-right intracardiac shunt (detects shunts with Qp/Qs≥1.01:1). and oximetry and indicator dilution techniques were used to quantitate the magnitude of shunting. Of the 20 patients, 11 had evidence of left-to-right shunting by hydrogen appearance times. Compared to catheterization, MRI correctly identified the 9 patients without intracardiac shunts and the 3 patients (2 PFO and 1 ASD) with insignificant shunts (Qp/Qs<1.5:1), as well as the 8 patients (6 ASD, 1 VSD, 1 PDA) with significant intracardiac shunts (Qp/Qs>1.5:1). In conclusion, velocity-encoded, phase-difference, cine magnetic resonance imaging can reliably determine the presence or absence of a significant left-to-right intracardiac shunt.