BACKGROUND There is convincing evidence of an association between the QT interval on electrocardiograms and general mortality. However, results are inconclusive regarding the extent to which this association depends on ventricular mass and size. METHODS Data were obtained from the prospective, population-based CARLA study, with a mean follow-up of 8.8 years, after exclusion of subjects with atrial fibrillation (919 men, 797 women aged 45-83 years remained eligible). Echocardiographic parameters were left ventricular mass index, left ventricular diastolic dimension index, diastolic interventricular septum thickness, diastolic left ventricular posterior wall and the relative left ventricular wall thickness. Heart rate-corrected QT interval (QTc) was measured with standard 12-lead electrocardiograms using the MEANS algorithm. The association between QTc and survival was modelled using Cox-regression models (crude- and covariate-adjusted). Values were standardized by dividing the QTc by the standard deviation. The association between QTc and survival was assessed in terms of tertiles of echocardiographic parameters. RESULTS In covariate-adjusted models, QTc was associated with general mortality (hazard ratio (HR): 1.19; 95% confidence interval (CI): 1.03, 1.38). Compared with higher tertiles, subjects in the lowest tertile of left ventricular mass index (HR=1.73, 95% CI: 1.26, 2.36) showed the strongest association with general mortality, which was also true for the lowest tertile of diastolic left ventricular posterior wall thickness (HR=1.49, 95% CI: 1.10, 2.02). CONCLUSION In the general population, the association between QTc and general mortality is strongest in subjects with low left ventricular mass index and diastolic thickness of the left ventricular posterior wall, thus the prognostic value of QTc needs to be interpreted with regard to these echocardiographic parameters.