Purpose: To prospectively determine the diagnostic performance of resting myocardial magnetic resonance imaging at 3 T for depiction of clinically relevant coronary artery stenosis (≥ 70% diameter) in patients with suspected or known to have coronary artery disease, using conventional coronary angiography as the reference standard and comparing results with SPECT findings.Materials and Methods: Cine MR imaging, first-pass perfusional images and delayed enhancement of the entire left ventricle were acquired in 31 patients at rest. Stress SPECT scan and selective coronary angiography were performed in all patients. Evaluation was performed on a per territory basis.Results: The overall sensitivity and specificity of MR imaging for depicting coronary artery disease was 83% (44 of 53 diseased territories) and 92% respectively. The overall sensitivity and specificity of SPECT for depicting coronary artery disease was 77% (41 of 53 diseased territories) and 95% respectively. The sensitivity of SPECT was higher than MR imaging for depiction of one vessel disease, yet cardiac MR showed higher sensitivity for depiction of double and triple vessel diseases. The sensitivity of cardiac MR and SPECT for detection of myocardial ischemic tissue was 68.9% and 72% respectively. The sensitivity of SPECT in detection of subendocardial scarring and scarred (non viable) myocardial segments was 14% and 54% respectively as referred to cardiac MR viability findings. Conclusion: Resting cardiac MR could be valuable in detection of significant coronary artery disease with relatively higher sensitivity values compared to stress SPECT study. In addition, it shows high accuracy in detection and delineation of myocardial scarring, particularly the subendocardial scarring.