Heart rate variability (HRV) patterns reflect the changing sympathetic and parasympathetic modulation of the autonomic nervous system. While overall and circadian heart rate (HR) and HRV are well characterized by traditional measures, ultradian cycles of HR and HRV are not. We have developed a method for capturing these rhythms during sleep and have now applied it to 24-hour recordings. METHODS: HR/HRV for each 2-min was calculated using normal-to-normal (NN) interbeat intervals from 24-hour Holter recordings in 10 healthy subjects, aged 26 +/- 2 yrs, 5M, 5F. HR, the standard deviation of NN intervals (SDNN2), high frequency power (HF) and the LF (low frequency power)/HF ratio were plotted. A curve-fitting algorithm, developed in MatLab, identified cyclic patterns of HR/HRV and extracted parameters to characterize them. Values were compared to those obtained in nighttime-only recordings in a set of 113 subjects, aged 58 +/- 10 yrs, 65M, 48F. RESULTS: Cyclic ultradian cycles were observed for each HR/HRV index. They had variable correspondences with each other and none could be considered surrogates. Although the number of cycles over 24 hours was greater, the mean cycle duration/number of cycles per hour was similar in both sets of recordings. CONCLUSIONS: Each HR/HRV parameter has its own rhythm, and the correspondence between these rhythms varies greatly across subjects. Although further studies are needed, it appears that there are intrinsic rhythms of autonomic modulation of HR on an scale of about 50 mins that persist during both the day and nighttimes. Quantification of ultradian patterns of HRV from 24-hour recordings is feasible and could provide new insights into autonomic physiology.