We have recently demonstrated that the overnight profiles of cardiac interbeat autocorrelation coefficient of R-R intervals (rRR) calculated at 1-min intervals are related to the changes in sleep electroencephalographic (EEG) mean frequency, which reflect depth of sleep. Other quantitative measures of the Poincaré plots, i.e., the standard deviation of normal R-R intervals (SDNN) and the root mean square difference among successive R-R normal intervals (RMSSD), are commonly used to evaluate heart rate variability. The present study was designed to compare the nocturnal profiles of rRR, SDNN, and RMSSD with the R-R spectral power components: high-frequency (HF) power, reflecting parasympathetic activity; low-frequency (LF) power, reflecting a predominance of sympathetic activity with a parasympathetic component; and the LF-to-HF ratio (LF/HF), regarded as an index of sympathovagal balance. rRR, SDNN, RMSSD, and the spectral power components were calculated every 5 min during sleep in 15 healthy subjects. The overnight profiles of rRR and LF/HF showed coordinate variations with highly significant correlation coefficients (P < 0.001 in all subjects). SDNN correlated with LF power (P < 0.001), and RMSSD correlated with HF power (P < 0.001). The overnight profiles of rRR and EEG mean frequency were found to be closely related with highly cross-correlated coefficients (P < 0. 001). SDNN and EEG mean frequency were also highly cross correlated (P < 0.001 in all subjects but 1). No systematic relationship was found between RMSSD and EEG mean frequency. In conclusion, rRR appears to be a new tool for evaluating the dynamic beat-to-beat interval behavior and the sympathovagal balance continuously during sleep. This nonlinear method may provide new insight into autonomic disorders.