The aim of this study was to examine whether changes in heart rate variability (HRV) can predict arrhythmias in children who have undergone the Fontan procedure. The study included 15 children with total cavopulmonary connection. All examinations included echocardiography and 24-h ambulatory electrocardiogram with power spectral analysis of HRV and Poincare plots (plots of each R-R interval as a function of the previous R-R interval). Six patients developed supraventricular tachycardia (four or more consecutive supraventricular beats). One patient was excluded from the study due to the development of bradycardia, necessitating placement of a pacemaker. Compared to the arrhythmia group, the patients without arrhythmias (n = 8) exhibited a significant difference in the standard deviation of instantaneous beat-to-beat R-R interval variability (p = 0.02). Poincare plots of patients without arrhythmias showed a typical torpedo- or comet-shaped pattern, whereas the arrhythmia group showed a more complex pattern. Although this study examined only a few subjects, needing confirmation by larger studies, it does indicate that patients who develop arrhythmias after receiving the Fontan procedure show a different Poincare pattern compared to the nonarrhythmic group-even before the arrhythmia can be detected using the conventional Holter procedure. Routine follow-up using the Holter procedure and Poincare plot analysis could help detect early arrhythmias.