Heart rate variability (HRV) is the output of multiple physiological control mechanisms that occurs over a wide range of complex time scales. The multiscale entropy (MSE) of RR interval time series is a well established complexity based nonlinear technique, to analyze HRV at different time-scales using sample entropy (SampEn). Determination of SampEn requires a priori determination of two parameters; pattern length (m) to be compared and tolerance threshold value (r) to accept the similarity between the patterns. This study aims to investigate the applicability of previously recommended value of r to be 0.1–0.2 times the standard deviation of time series for MSE analysis. MSE at higher scales are found to be inappropriate with recommended range of r. Therefore to find maximum MSE at higher time scales, calculation of MSE with different possible values of r need to considered. But this approach is very time consuming and arduous. Further, MSE is found to be decreased significantly in function of time scales.