Background: QRS duration and left bundle branch block (LBBB) morphology are used to select patients for cardiac resynchronization therapy (CRT). We investigated whether the area of the QRS complex (QRS(AREA)) on the 3-dimensional vectorcardiogram (VCG) can improve patient selection. Methods: VCG (Frank orthogonal lead system) was recorded prior to CRT device implantation in 81 consecutive patients. VCG parameters, including QRSAREA, were assessed, and compared to QRS duration and morphology. Three LBBB definitions were used, differing in requirement of mid-QRS notching. Responders to CRT (CRT-R) were defined as patients with 15% reduction in left ventricular end systolic volume after 6 months of CRT. Results: Fifty-seven patients (70%) were CRT-R. QRSAREA was larger in CRT-R than in CRT nonresponders (140 +/- 42 vs 100 +/- 40 mu Vs, p < 0.001) and predicted CRT response better than QRS duration (AUC 0.78 vs 0.62, p = 0.030). With a 98 mu Ns cutoff value, QRSAREA identified CRT-R with an odds ratio (OR) of 10.2 and a 95% confidence interval (CI) of 3.4 to 31.1. This OR was higher than that for QRS duration >156 ms (OR = 2.5; 95% CI 0.9 to 6.6), conventional LBBB classification (OR = 5.5; 95% CI 0.9 to 32.4) or LBBB classification according to American guidelines (OR = 4.5; 95% CI 1.6 to 12.6) or Strauss (OR = 10.0; 95% CI 3.2 to 31.1). Conclusion: QRS(AREA) is an objective electrophysiological predictor of CRT response that performs at least as good as the most refined definition of LBBB. Condensed abstract: In 81 candidates for cardiac resynchronization therapy (CRT) we measured the area of the QRS complex (QRS(AREA)) using 3-dimensional vectorcardiography. QRS(AREA) was larger in echocardiographic responders than in non-responders and predicted CRT response better than QRS duration and than simple LBBB criteria. ORSAREA is a promising electrophysiological predictor of CRT response. (C) 2015 Elsevier Inc. All rights reserved.