Autonomic nervous function was evaluated by means of power spectral analysis of heart rate variability in hospitalized dipper (n = 31) and non-dipper (n = 31) essential hypertensive subjects. Twenty-four-hour blood pressure (BP) measurement was performed by the cuff-oscillometric method to evaluate the nocturnal decrease of BP. The non-dipper subjects were defined as those whose nocturnal decrease of systolic BP was < 10% of daytime BP. Power spectral analysis of RR interval was performed from Holter ECG every 10 minutes by the maximum entropy method to obtain the low-frequency band (LFB, 0.04 to 0.15 Hz), which is an index of both parasympathetic and sympathetic nervous activities, and the high frequency band (HFB, 0.15 to 0.4 Hz), which reflects parasympathetic nervous activity. LFB and HFB were averaged every hour to obtain hourly LFB and HFB values. Total LFB and total HFB were calculated as the mean values of 24 hourly averaged LFBs and HFBs. Both LFB and HFB were significantly lower in non-dipper hypertensives than in dipper subjects throughout the day. In dipper hypertensives, LFB showed a nocturnal decrease, whereas HFB was significantly increased during the nighttime. However, these diurnal changes in LFB and HFB were significantly blunted in non-dipper subjects. These findings indicate that non-dipper hypertensive subjects were characterized with a decreased physiological circadian fluctuation on autonomic functions compared with dipper subjects. This alteration in the autonomic nervous function may explain the non-dipper phenomenon in essential hypertension.