The challenge for therapies targeting perfusion abnormalities is to identify and evaluate the region of interest. The aim of this study was to compare rest and stress myocardial perfusion measured by cardiac multi-detector computed tomography (MDCT) and cardiac magnetic resonance (CMR) imaging in patients with invasive coronary angiography demonstrated occluded vessels. Twenty-four patients with refractory angina due to occluded coronary arteries underwent perfusion imaging obtained by 320-MDCT scanner and 1.5 T MR scanner. Rest and adenosine stress images were obtained and interpreted using the modified 17-segment American Heart Association model. For the qualitative analysis, each segment was graded according to the following scoring system: 0 = no defect, 1 = hypoperfusion transmural extent <1/3, 2 = 1/3-1/2, 3 = >1/2, and 4 = infarct stigmata. In the semiquantitative analysis the perfusion was either scored 0 (normal) or 1 (abnormal). The summed rest and stress scores were calculated. MDCT and CMR had a high probability to identify perfusion defects. An excellent correlation between MDCT and CMR summed rest (r = 0.916) and stress scores (r = 0.915) was found. The interobserver reproducibility was high for MDCT and CMR images. The qualitative and semiquantitative MDCT against CMR analysis of rest and stress images showed high concordance to detect perfusion defects per vascular territory and on a per myocardial segment basis. 320-MDCT and CMR perfusion imaging can be used clinically to identify myocardial perfusion defects and potentially evaluate the effect of therapy targeting perfusion abnormalities.