Heart rate variability, as determined from 24-hour Holter recordings, represents a noninvasive parameter for studying the autonomic control of the heart. It decreases with certain disease states characterized by autonomic dysfunction such as congestive heart failure. No study in healthy or cardiac children has been performed to determine the correlations between and within time and frequency domain indices of heart rate variability. We examined five time domain (SDNN, SDNNi, SDANNi, rMSSD and pNN50) and five frequency domain measures (ULF, VLF, LF, HF and balance LF/HF) in 200 healthy children and 200 children with congenital heart disease, aged 3 days to 14 years. All measures were significantly correlated with each other. However, the strength of correlation varied greatly. Our data show that variables strongly dependent on vagal tone (rMSSD, pNN50 and HF) were highly correlated (r value > 0.90), as well as SDNN and SDANNi. We conclude that certain time and frequency domain indices correlate so strongly with each other that they can act as surrogates for each other.