To assess the combined influence of blood pressure and body mass index (BMI) on death from all-cause and circulatory diseases, a retrospective cohort study was performed in a community of Kyoto from 1978 to 1991. Between 1978 and 1982, 1,629 people (650 men and 979 women) aged 40-69 years, representing 76.6% of this age group in the community, were screened for entry into this study. A total of 134 deaths were observed and death by circulatory diseases was 36.6% of the all-cause deaths. The survival rate, calculated by the Kaplan-Meier method, was higher among normotensives [systolic blood pressure (SBP) < 140 mmHg and diastolic blood pressure (DBP) < 90] compared to borderline hypertensives (SBP = 140-159 and/or DBP = 90-94) and hypertensives (SBP > = 160 or DBP > = 95, or using antihypertensive medication) for subjects aged 50-69 years both males and females. The hazard ratio of all-cause and circulatory disease mortality for sex, age, blood pressure and BMI as covariables was estimated from the Cox proportional hazard model. The hazard ratio (95% confidence interval) was greater for circulatory disease mortality [2.76 (1.37-5.75) for borderline hypertensive and 2.53 (1.21-5.26) for hypertensive] than for all-cause mortality [1.86 (1.22-2.84) for borderline hypertensive and 2.03 (1.32-3.10) for hypertensive]. These results indicate that hypertension increases risk of circulatory disease mortality and decreases survival rate. While the effect of obesity was not as strong for circulatory disease mortality as blood pressure, an increase in hazard ratio with increasing BMI was suggested in this study.