SummaryAims: The ECG phenomenon called localised right precordial QRS prolongation in the diagnosis of arrhythmogenic right ventricular dysplasia-cardiomyopathy defined as a QRS duration in V1, V2 or V3110ms has raised much interest in QRS duration in healthy volunteers. The purpose of the present study was to analyse QRS durations and special ratios in a large ECG database of healthy volunteers in order to obtain more information about the distribution of QRS duration in different leads. Methods and results: In a database of standard 12-lead ECGs of 1560 healthy volunteers (888 males and 672 females) mean, standard deviation and the 96% range of QRS duration for each lead were determined by computer using the well-established Glasgow Program. The values of a ratio of QRS duration in leads (V2+V3)/(V4+V5), of a ratio of QRS duration in leads V1/I and the presence of QRS duration in right precordial leads V1, V2 or V3110ms were derived. Only in 3.5% of males, but in no females, the QRS duration in V1, V2 or V3 was 110ms. If 1.2 was used as the critical normal value of QRS duration ratio in (V2+V3)/(V4+V5) specificity was 90.8% for males and only 79.5% for females. Conclusion: QRS duration 110ms in the right precordial lead V1, V2 or V3 seems to be the best marker for discriminating healthy volunteers from patients with the phenomenon of localised right precordial QRS prolongation. The comparison of QRS duration in right and left precordial leads lacks specificity predominantly in females if a cut-off value of 1.2 is used.