Hyperinsulinemia has been closely associated with hypertension in several epidemiological studies, but little is known about this condition in the elderly. The authors studied the relationship between serum insulin and blood pressure levels, and the prevalence of hypertension according to insulin levels in the elderly in a Japanese rural community, Hisayama. In 1988, 75 g oral glucose tolerance test was performed on 426 male and 567 female Hisayama residents aged 60 to 79 years. Fasting and 2 hr serum insulin values were measured by radioimmunoassay. In the subjects, excluding those receiving antihypertensive drugs, the sum of fasting and 2 hr postload insulin (sigma IRI) significantly correlated with systolic (r = 0.15 and 0.25 for males and females, respectively) and diastolic blood pressure (r = 0.20, 0.16). In multiple regression analyses, the correlation with systolic blood pressure remained significant in females after controlling for age, body mass index, alcohol intake, smoking habits, serum total cholesterol, HDL-cholesterol, triglycerides, and fasting plasma glucose. In contrast, sigma IRI did not remain as an independent variable relevant to blood pressure among males. The age-adjusted prevalence rates of hypertension (> or = 160/95 mmHg or receiving drug treatment) significantly increased with increasing quartiles of sigma IRI in females. The prevalence of hypertension which was not treated with drugs also increased significantly with increasing sigma IRI in females. However, no significant association was observed in males. In conclusion, the present study suggests that hyperinsulinemia can be related to hypertension in the female elderly in the general population of Japan.