Purpose: Cardiac resynchronization therapy (CRT) is an established treatment for patients with advanced heart failure. However, determinants of response to CRT remain elusive. The aim of the study was to assess the value of ECG parameters to predict super-response in patients treated with CRT. Methods: A 12-lead surface ECG was recorded at baseline and immediately after implantation of CRT in 101 patients with advance heart failure and sinus rhythm (age 63.2 ± 10.9; NYHA class II-IV; 66 (65.3%) male; 37 (36.6%) ischaemic aetiology). Baseline ECG parameters (native QRS duration, bundle branch morphology, axis, PR interval, and QT/QTc duration) and post-implant paced QRS duration were analysed; relative change in QRS duration was also calculated. Patients with a decrease in left ventricular end-systolic volume >30% after 6 months were classified as super-responders. Results: Thirty-two patients (31.7%) were classified as super-responders. There were no significant differences in baseline ECG parameters between super-responders and other CRT patients. Post-implant QRS duration was shorter in super-responders (147.6 ± 21.7 vs.162.4 ± 28.1; p=0.005). Additionally, in super-responders significant QRS reduction (p=0.002) and relative change in QRS duration (p=0.005) was observed. After receiver-operating curve analysis relative change of QRS duration > 4.5% had an 81% sensitivity and 58% specificity for predicting super-responders. In a multivariate analysis relative change in QRS duration remained independent predictor of super-response. Conclusion: Absolute post-implant QRS duration and change in QRS duration are the only ECG parameters associated with super-response in CRT.