A cross-sectional community survey using home blood pressure measurements was performed in northern Japan to estimate the prevalence of definite hypertension, white coat hypertension and the success of blood pressure control in patients receiving antihypertensive drugs. A total of 1334 subjects (mean age +/- SD, 53.8 +/- 17.3 years; 8-91 years) participated in the screening and home blood pressure measurement program. They measured blood pressure at home at least 3 times (mean measurement frequency, 20.8 +/- 8.3 times). Of these 1334 subjects, 314 (65.1 +/- 8.9 years) were taking drugs (treated group) while 1020 (50.3 +/- 17.8 years) were not (untreated group). The WHO criteria were used to categorize screening blood pressure. Criteria for diagnosis of hypertension by home blood pressure measurements were as follows: definitely hypertensive (systolic blood pressure > or = 144 mmHg and/or diastolic blood pressure > or = 89 mmHg) and normotensive (104 < systolic blood pressure < or = 131 mmHg and 60 < diastolic blood pressure < or = 79 mmHg). Of the 1018 subjects identified as normotensive on screening measurements, home measurements indicated that 73 (7.2%) were hypertensive and 765 (74.7%) were normotensive or lower. Of the 112 subjects identified as hypertensive on screening measurements, home measurements showed that 42 (37.5%) were hypertensive and 30 (26.8%) were normotensive or lower. Of the 314 treated subjects, 45 (14.3%) were identified as hypertensive by screening measurements and 88 (28.0%) as hypertensive by home measurements. Only 20 (44.4%) of the former 45 subjects were also defined as definitely hypertensive by home measurements. Of the 1020 untreated subjects, 67 (6.6%) were hypertensive by screening measurements and 84 (8.2%) by home measurements. Only 22 (32.8%) of the former 67 subjects were classified as hypertensive by home measurements. Of the 67 untreated subjects identified as hypertensive by screening measurements, 20 (29.9%) were normotensive or lower by home measurements, suggesting that these subjects were "white coat" hypertensives. The study first confirmed based on the large community data that there are large discrepancies between screening (casual) blood pressure and home blood pressure measurements for recognition of hypertension and normotension. Determination of blood pressure levels by home blood pressure measurements may predict prognosis of hypertension differently from that by screening blood pressure measurements. Further prospectively study is needed to validate the prognostic value of home blood pressure measurements.