Background: A close relationship between alcohol consumption and hypertension has been established, but the effect of light to moderate alcohol consumption on blood pressure (BP) remains unclear. Methods: A total of 5275 Japanese male office workers aged 23 to 59 years partook in a survey. Subjects were classified as nondrinkers or current drinkers who averaged less than 12, 12 to 22, 23 to 45, or greater than or equal to46 g/day of ethanol. BPs were measured at annual health examinations from May 1996 through May 2000. A total of 3784 hypertension-free (systolic BP <140 mm Hg, diastolic BP <90 mm Hg, no medication for hypertension, and no history of hypertension) men were observed for 4 years. Men in whom hypertension (systolic BP greater than or equal to140 mm Hg, diastolic BP greater than or equal to90 mm Hg, or both or receipt of antihypertensive medication) was found during repeated surveys were defined as incident cases of hypertension. Results: After controlling for potential predictors of hypertension, systolic and diastolic BP levels and the incidence of hypertension were lowest in nondrinkers in all three age ranges (23-35, 36-47, and 45-59 years) and increased in a dose-dependent manner as alcohol consumption increased. For individuals aged 23 to 35 years, systolic and diastolic BP levels and the incidence of hypertension were significantly higher among those who drank an average of greater than or equal to23 g/day of ethanol than among nondrinkers. For those aged 36 to 59, the diastolic BP level was significantly higher among those who drank any alcohol at all than among nondrinkers, and the systolic BP level and the incidence of hypertension were significantly higher among those who drank an average of greater than or equal to12 g/day of ethanol than among nondrinkers. Conclusions: Light to moderate alcohol consumption seems to have an important influence on BP in both young and middle-aged Japanese men.