Background: Cardiac resynchronization therapy (CRT) is an established therapy in heart failure. However, one third of patients are non-responders, accordingly, proper selection of CRT eligible patients are essential.Aims: 1) To test their ability to predict response to CRT in patients with ischemic and non-ischemic cardiomyopathy. 2) To compare TomTec and QLAB software packages for the three-dimensional echocardiographic (3DE) assessment of left ventricular (LV) dyssynchrony. Methods and results: A total of 140 heart failure patients with the LVEF ≤35% and 60 healthy volunteers underwent 3DE. A subgroup of 60 patients underwent CRT and were evaluated before and 6-12 months after implantation. The systolic dyssynchrony index (SDI) for all 16 LV segments was measured with both software packages and compared using Pearson’s correlation and Bland–Altman analysis. Measurements of SDI were significantly higher using TomTec compared with QLAB in both patients (10.9 ± 3.8 vs. 9.7 ± 3.9, P<0.001) and healthy volunteers (4.1 ± 0.8 vs. 2.4 ± 1, P <0.001), with large biases and wide limits of agreement. A moderate correlation (r= 0.65, P <0.001) was observed between both software packages in patients while their inter-observer and intra-observer reliability were good. Of the 60 patients undergoing CRT, reverse remodeling as a measure of response was observed in 41 patients (68%). The optimal SDI cut-off value to predict response to CRT was higher for TomTec than for QLAB (8.8 vs.7.3%, P <0.001) and demonstrated better sensitivity and specificity (93 and 61%, respectively) compared with QLAB (88 and 33%, respectively). Response prediction in patients with non-ischemic cardiomyopathy was excellent with a sensitivity and specificity of 95 and 100% for TomTec and 70 and 83% for QLAB.Conclusion: Dyssynchrony assessment with 3DE for the prediction of response to CRT seems particularly useful in patients with non-ischemic cardiomyopathy. Different 3DE software packages for the assessment of mechanical dyssynchrony should not be used interchangeably until better software standardization is achieved.