BACKGROUND The analysis of the circadian rhythm of heart rate variability (HRV) represents a relevant physiological tool to assess the vagal system. However, the influence of age (mostly >75 years) on HRV is not widely known. OBJECTIVE The aim of this study was to evaluate the influence of age on HRV, particularly in the elderly and to identify a model of this relationship. The study was carried out by examining linear and nonlinear parameters extracted from HRV, including individuals over 75 years for which there is no research available. METHODS Data from 140 healthy subjects were sex matched and divided into young (young group: 15-39 years old), adult (adult group: 40-64 years old) and senior (senior group: 65-90 years old) groups. 24-h Holter monitoring was used and several HRV parameters were extracted from time, and spectral and nonlinear analyses were examined. RESULTS Time-domain parameters, mainly standard deviation of the NN interval (SDNN) and number of successive differences of intervals which differ by more than 50 ms, presented significant differences between the young group and the other two groups during the 24-h period, while normalized spectral parameters (LFn, HFn and low frequency/high frequency), as well as nonlinear parameters, mainly β exponent and fractal dimension, showed significant difference between the senior group and the other two groups. All these parameters showed a similar circadian rhythm with significant differences between the mean day and night values, especially in young and adult group cohorts. Moreover, a parabolic relationship between these parameters and age was highlighted with an opposite trend over about 60 years compared with younger people. CONCLUSIONS A progressive physiological autonomic imbalance is present in ageing. The inverse trend in the relation between HRV parameters and age found in the senior group could be mainly due to a faster fluctuation of RR. This should be considered when studying changes in the cardiac autonomic nervous control.