The assessment of regional volumes is an option for analysis of the response of LV segments to interventions such as revascularization or cell therapy. We sought to compare regional volumes from 3D-echocardiography (3DE) with cardiac magnetic resonance (CMR) over follow-up. CMR regional volumes were assessed at baseline and after one year follow-up in 30 unselected patients (28 men, 65 +/- 11 years) presenting for evaluation of cardiac function with previous infarction. 3DE images were also gathered over 4 cardiac cycles and measurements were performed off-line. CMR images were obtained using a 1.5 Tesla scanner and measured offline by method of landmarks and by centre of mass. Regional volumes were measured at end-diastole (rEDV) and end-systole (rESV) and the change in volume was compared for each over follow-up. There was good correlation between 3DE and both CMR methods at baseline and follow-up. Changes in rEDV with 3DE vs CMR(L) were comparable (0.11 +/- 3 ml vs 0.12 +/- 3 ml, p = 0.94), as was change in CMR(M) (0.26 +/- 2 ml, p = 0.69). However the change in regional volume by 3DE and CMR(L) correlated poorly (r = 0.03, p = 0.68), as did change in 3DE vs CMR(M) (r = 0.04, p = 0.65). Similarly, changes in rESV with 3DE and CMR(L) were similar (0.27 +/- 2 ml vs 0.36 +/- 2 ml, p = 0.70), as was change in CMR(M) (0.05 +/- 1 ml, p = 0.31). Again, correlations between rESV by 3DE vs CMR(L) were poor (r = 0.03, p = 0.72), as well as 3DE vs CMR(M) (r = 0.07, p = 0.40). Although global 3DE volumes compare well with CMR volumes, new developments in image quality and automated software will be needed before changes in regional volumes can be reliably followed with 3DE.