A 1.5 year community-based hypertension education program was conducted and evaluated for the feasibility and effectiveness as a program for primary prevention of cardiovascular disease. The subjects were determined to be hypertensive from cardiovascular examinations and composed of 104 untreated men and women aged 35-69 whose blood pressure levels were between 140 and 179 mmHg for systolic and/or between 90 and 109 mmHg for diastolic and who had no evidence of hypertensive end organ effects in the 1990 examination. They were randomly assigned to either an intensive education (IE) group (n = 53) or a usual education (UE) group (n = 51) and invited to education classes six months later. The IE consisted of four education classes in the first six months and another 4 classes in the next 12 months. For the UE, one class in the first 6 months and another class in the next 12 months were held. Health education was focused on reduction of sodium intake, a 30 minutes of walking, and if necessary, reduction of alcohol and sugar intake. Mean (SD) values of systolic blood pressure at the start of the education program were 144.8 (11.7) mmHg in IE and 144.5 (12.2) mmHg in UE. The respective diastolic means (SD) were 80.6 (7.8) mmHg and 80.3 (8.6) mmHg. The proportion of persons who started to receive antihypertensive medication was 8-9% in both groups at six months, and 13% in IE and 29% in UE at 1.5 years. IE showed a 6 mmHg greater reduction in mean systolic blood pressure than UE at both six months and 1.5 years. No significant difference was found in diastolic blood pressure levels over time between the two groups. Dietary improvements related to reduction of salt intake were more evident in IE than in UE. There was no significant changes in means and distributions of relative weight index and alcohol intake in either group. Subgroup analysis according to antihypertensive medication use indicated that reduction of systolic blood pressure levels was significantly larger in IE than in UE for both no medication and medication subgroups at 6 months and 1.5 years. While it was difficult to evaluate an effect of the program at 1.5 years because of the difference in the proportion of medication use between IE and UE, the results at 6 months suggest that this community-based program was effective in reducing systolic blood pressure levels.