Analysis of heart rate variability (HRV) is performed through interbeat interval time series derived from either electrocardiographic or arterial pressure (AP) recordings. However, little attention has been given to the reliability of calculating the time series from different sources, ie electrocardiogram (ECG) or pulse intervals (PI). Therefore, the present study aimed to evaluate the correlation between interbeat interval time series obtained from RR, inter-systolic (SS) and inter-diastolic (DD) intervals, as well as their impact on indices of HRV calculated from series of RR or PI. Conscious rats previously instrumented with subcutaneous electrodes and a catheter into the femoral artery were subjected to simultaneous ECG and AP recording for 5 min. Correlation and Bland–Altman plots between RR and PI were evaluated. Moreover, HRV was analyzed in time (mean cardiac interval, SDNN and RMSSD) and frequency domain (power in LF and HF spectral bands) as well as by nonlinear approaches (symbolic dynamics and sample entropy). First, RR showed a stronger correlation with PI calculated by DD than SS. Second, most HRV indices showed similar results when calculated with RR or DD series, but not with SS series. Considering RR interval as the gold standard for the calculation of cardiac cycle, when using PI inter diastolic intervals are the better choice to study HRV. These findings are quite relevant, especially when AP recording is used for HRV analysis.