We analyzed by means of autoregressive spectral analysis the spontaneous beat-to-beat heart rate variability (HRV) of quadriplegic and paraplegic male subjects at rest in the supine position. In agreement with our previous study, in nine of 15 quadriplegic patients only the high-frequency (HF: center frequency = respiratory frequency) component (a marker of vagal modulation of heart rate) was observed. In contrast, in six of the quadriplegic patients both the HF component and the low-frequency (LF: center frequency at approx. 0.1 Hz, 0.03–0.15 Hz in this study) component (a marker of sympathetic and vagal modulation of heart rate) were observed. However, in six quadriplegic patients who presented the LF component, ( i) the center frequency of the LF component was lower than that in 10 healthy, sedentary, age-matched males (control I) ( P < 0.01), ( ii) the power of the HF component was smaller than that in the control-I group ( P < 0.01) and ( iii) the LF/HF power ratio (an index of sympathovagal balance) was larger than that in the control-I group ( P < 0.05). On the other hand, in nine paraplegic patients with an intact 1st–4th thoracic spinal cord, from which the cardiac sympathetic nerves originate, the total power, the power of the LF component and that of the HF component were smaller than those in nine healthy, sedentary, age-matched males (control II) ( P < 0.05, P < 0.01 and P < 0.01, respectively). These results suggest that ( i) in quadriplegic patients who presented the LF component, the physiological mechanism of the LF component may be different from that in the control-I subjects (e.g., contribution of spinal sympathetic nervous system), ( ii) in paraplegic patients having both the intact cardiac sympathetic nerves and the intact vagal cardiac nerves, the dysfunction of the sympathetic nerves to the vessels below the level of the lesion and/or the compensatory vagal suppression may contribute in part to the low HRV and ( iii) it is possible that in some clinical pathophysiological conditions the dysfunction of the sympathetic nerves to the vessels and/or the compensatory vagal suppression may affect the HRV. Therefore, one must give careful consideration to the definition (e.g., center frequency) and interpretation of the LF component and to the interpretation of results when analyzing the HRV in humans by means of power spectral analysis.