OBJECTIVE--To study QT dispersion in left ventricular hypertrophy and chronic heart failure and to determine the relation to ventricular arrhythmias. SETTING--Investigational laboratory of a tertiary referral centre. STUDY DESIGN--Patients with left ventricular hypertrophy and normal systolic function (n = 14) and patients with chronic heart failure (n = 18) were matched with controls (n = 17). The QT dispersion was examined in relation to abnormalities in resting mechanical and autonomic function and to the findings of 24 hour Holter monitoring. MAIN OUTCOME MEASURES--QT dispersion is the difference between the maximum and the minimum QT values from the 12 lead electrocardiogram. Mean(SD) QT dispersion from the 10 lead electrocardiogram was also examined once the 12 lead minimum and maximum values had been removed. The QT distribution is the curve describing the distance from the mean for all QT intervals (ms). RESULTS--All measures of QT dispersion were increased significantly in left ventricular hypertrophy and tended to increase in those with heart failure. The QT distribution was abnormal in both heart failure and left ventricular hypertrophy. There was no relation between the degree of change in QT dispersion and the incidence of ventricular arrhythmia on 24 hour Holter monitoring. Also there was no relation between QT dispersion and autonomic or mechanical abnormalities. The QT dispersion was related to QRS duration. CONCLUSION--Though QT dispersion and distribution are abnormal in left ventricular hypertrophy these findings do not support the hypothesis that QT dispersion reflects arrhythmic risk in either hypertrophy or heart failure.