We investigated the effects of alcohol restriction on ambulatory blood pressure (BP), heart rate, and heart rate variability in 33 Japanese male volunteers (37 +/- 1 years, mean +/- SE), who were all habitual drinkers. Subjects were told either to keep their usual drinking habits for 3 weeks (usual alcohol period), or to reduce alcohol intake by at least half of their usual drinking amount (reduced alcohol period). The ambulatory BP, heart rate, and electrocardiographic R-R intervals were measured during a 24-h period with a portable recorder on the last day of each period. A power spectral analysis of R-R intervals was performed to obtain the low-frequency (LF) and high-frequency (HF) components. The percentage of differences between adjacent normal R-R intervals >50 msec (pNN50) was also calculated. The amount of ethanol intake was significantly reduced from 70 +/- 5 mL/day in the usual alcohol period to 19 +/- 3 mL/day in the reduced alcohol period (P < .0001). The daytime systolic BP was significantly lower in the reduced alcohol period than in the usual alcohol period by 4 +/- 1 mm Hg (P < .05). The daytime and nighttime heart rate was significantly lower in the reduced alcohol period than in the usual alcohol (P < .001 for each). The pNN50 and the HF component were significantly higher in the reduced alcohol period than in the usual alcohol period (P < .0001 for each). The LF/HF ratio was significantly lower in the reduced period than in the usual period (P < .01). These results demonstrate that 3-week alcohol restriction produced reductions in ambulatory systolic BP, heart rate, and the index of sympathovagal balance, and augmentations of parasympathetic indices of heart rate variability in Japanese male drinkers.