OBJECTIVE To investigate the effect of carvedilol on electrocardiographic parameters in children with congestive heart failure. METHODS 18 children with heart failure (aged 2 months-17 years) were treated with carvedilol (initially 0.09 mg/kg/day, slowly increased up to 0.7 mg/kg/day) in addition to conventional therapy with digoxin, ACE inhibitors and diuretics. Twelve-lead rest electrocardiograms (ECGs) and echocardiography were performed in 16 patients with sinus rhythm at baseline and after 1, 2, 4 (n = 14) and 6 months (n = 14) of therapy. ECGs were analysed for heart rate, QT duration and QT dispersion. Echocardiography was performed for analysis of ejection fraction. RESULTS After 6 months of therapy the mean ejection fraction increased from 37% to 55% (p < 0.05) and mean heart rate decreased by 14% (p < 0.05). Mean QT duration calculated by Bazett's formula (QT(B)) and Fridericia's formula (QT(F)) decreased from 428 msec (372-507 msec) to 387 msec (323-440 msec [QT(B)]; p < 0.05) and from 381 msec (315-466 msec) to 355 msec (309-435 msec [QT(F)]; p < 0.05) following therapy with carvedilol. In contrast, mean QT dispersion did not change significantly (18 msec; 10-40 msec before to 12 msec; 5-20 msec; p > 0.05). CONCLUSIONS In conclusion, carvedilol treatment reduced QT duration but not QT dispersion in paediatric patients with heart failure. The decrease in QT duration reflects stabilisation of the action potential, and this may contribute to the improved prognosis in these patients.