We compare the diagnostic accuracy of multidetector row computed tomography (MDCT) to cardiac magnetic resonance imaging (CMR) for evaluating global left ventricular function. We systematically searched PubMed, CINAHL, Cochrane CENTRAL, Scopus, and the Web of Science databases for studies published between 1966 to January 2013 that compared left ventricle (LV) volumes, ejection fraction (EF) and LV mass measured by MDCT and CMR. We performed meta-analyses and used random-effects model with inverse variance weighting test to determine the overall bias and limits of agreement of LV end-diastolic volume, end-systolic volume, stroke volume, and EF measured by MDCT and CMR. Furthermore, subgroup analyses were performed to compare 16-slice and 64-slice MDCT with CMR. Two study authors independently reviewed the 90 articles originally identified and selected 27 studies (n = 831) for analysis. Excellent correlation and a linear relation were seen between MDCT and CMR for LV end-diastolic volume (r = 0.93; p <0.001), LV end-systolic volume (r = 0.95; p <0.001), LV stroke volume (r = 0.85; p <0.001), LV ejection fraction (r = 0.93; p <0.001), and LV mass (r = 0.86; p <0.001). Subgroup analyses showed strong positive correlations for both 16- and 64-slice MDCT. In conclusion, although not the first-line test for LV function assessment in most patients, when appropriate, retrospectively gated MDCT provides an accurate and valid assessment of LV function compared with CMR.