Measurements of the Q-T interval are less reliable in children than in adults. Identification of superior diagnostic tools is warranted. This study aimed to investigate whether a vectorcardiogram (VCG) recorded from three orthogonal leads (X, Y, Z) according to Frank is superior to a 12-lead electrocardiogram (ECG) in providing a correct long Q-T syndrome (LQTS) diagnosis in children. This LQTS group consisted of 35 genetically confirmed carriers of mutations in the KCNQ1 (n = 29) and KCNH2 (n = 6) genes. The control group consisted of 35 age- and gender-matched healthy children. The mean age was 7 years in the LQTS group and 6.7 years in the control group (range, 0.5-16 years). The corrected Q-T interval (QT(c)) was measured manually (QT(man)) by one author (A.W.). The 12-lead ECG automatic measurements (QT(ECG)) and interpretation (QT(Interpret)) of QT(c) were performed with the Mac5000 (GE Medical System), and the VCG automatic measurements (QT(VCG)) were performed with the Mida1000, CoroNet (Ortivus AB, Sweden). By either method, a QT(c) longer than 440 ms was considered prolonged and indicative of LQTS. Of the 35 children with genetically confirmed LQTS, 30 (86 %) received a correct diagnosis using QT(VCG), 29 (82 %) using QT(man), 24 (69 %) using QT(ECG), and 17 (49 %) using QT(Interpret). Specificity was 0.80 for QT(VCG), 0.83 for QT(man), 0.77 for QT(ECG), and 0.83 for QT(Interpret). The VCG automatic measurement of QT(c) seems to be a better predictor of LQTS than automatic measurement and interpretation of 12-lead ECG.