Aim: The influence of location and extent of transmural scar and its relation with dyssynchrony in cardiac resynchronization therapy (CRT) was investigated as posterolateral scar tissue has been invoked as a cause of non-response to CRT. Methods and results: Fifty-seven patients eligible for CRT were assessed for transmural scar with gadolinium-enhanced MRI and for left ventricular (LV) dyssynchrony with tissue Doppler. After implant, both atrioventricular and interventricular pacing intervals were optimized. LV reverse remodeling was defined as 10% decrease in LV end-systolic volume after 3 months. Sixteen patients had transmural scar in the posterolateral (PL) area (LV lead location), 14 at a remote site (non-PL) and 27 patients had no scar. LV reverse remodeling was observed in respectively 25%, 64% and 89% (P = 0.0001). Univariate analyses showed a relation with LV dyssynchrony (P = 0.004) and with absence of PL scar (P = 0.04) but not with QRS duration and the extent of LV scar tissue. In multivariate analysis, only LV dyssynchrony (OR: 19.62; 95% CI: 2.5–151.9; P = 0.004) independently predicted LV reverse remodeling. Conclusion: In this study LV dyssynchrony remains the most important determinant of response to CRT, even in the presence of posterolateral scar provided atrioventricular and interventricular pacing intervals are optimized.