Heart rate variability (HRV) is associated with blood pressure levels; however, very few studies have correlated HRV to lifestyle in the general population. We investigated 1418 men and 2040 women aged 40-74 years and measured the HRV indices in the time and frequency domains using a 5-min R-R interval recording. Multiple linear regression analyses were used to estimate the association between HRV and blood pressure levels after adjustment for major confounders. HRV indices were not associated with systolic blood pressure levels in men, whereas in women, one-s.d. (1-s.d.) increment of s.d. of normal R-R intervals (SDNN) was associated with a 1.21-mm Hg decrease in systolic blood pressure after adjusting for several confounders (P<0.05). In addition, a 1-s.d. increment of SDNN corresponded with 1.00 and 1.10 mm Hg reductions in diastolic blood pressure in men and women, respectively (P<0.05). When stratified by the use or nonuse of antihypertensive medication, these inverse associations were more evident in the nonuser group. Furthermore, among men not using antihypertensive medication, reduced HRV was associated with increased systolic blood pressure levels in nondrinkers. The data suggest that HRV reflects diastolic blood pressure better than systolic blood pressure levels for both sexes and that alcohol intake strongly affects systolic blood pressure levels in men, which may have weakened the association with HRV.