To perform a head-to-head comparison between magnetic resonance imaging (MRI) and gated single-photon emission computed tomography (SPECT) for the evaluation of left ventricular (LV) function (LV ejection fraction [LVEF], LV volumes, and regional wall motion) in patients with ischemic cardiomyopathy, we studied 22 patients with chronic coronary artery disease and LV dysfunction. Multislice, multiphase echoplanar MRI was performed with Philips Gyroscan ACS-NT15. Image analysis was performed using the MASS software package to determine LV end-systolic volume, LV end-diastolic volume, and LVEF. The same parameters were calculated using quantitative gated SPECT software (QGS, Cedars-Sinai Medical Center). The different parameters were compared using linear regression, and correlation coefficients were calculated. Regional wall motion was also determined from both techniques, according to a 13-segment model and a 3-point scoring system (from 1 = normokinesia to 3 = akinesia or dyskinesia). A summed wall motion score was also calculated for MRI and gated SPECT. Good correlations were found between MRI and gated SPECT for all parameters: (1) summed wall motion scoreMRI versus summed wall motion scoreSPECT: y = 0.74x + 8.0, r = 0.88, p <0.01; (2) LV end-systolic volumeMRI versus LV end-systolic volumeSPECT: y = 0.94x - 12.3, r = 0.87, p <0.01; (3) LV end-diastolic volumeMRI versus LV end-diastolic volumeSPECT: y = 0.93x - 18.4, r = 0.84, p <0.01; and (4) LVEFMRI versus LVEFSPECT: y = 0.97x + 0.68, r = 0.90, p <0.01. For regional wall motion, an exact agreement of 83% was found, with a kappa statistic of 0.77 (95% confidence intervals 0.71 to 0.83, SE 0.03), indicating essentially excellent agreement. Thus, close and significant correlations were observed for assessment of LVEF, LV volumes, and regional wall motion by MRI and gated SPECT in patients with ischemic cardiomyopathy.