Objective: To determine the relationship between pulse pressure and mortality from all the causes: CVD and CVA. Methods: The cohort consisted of the beneficiaries from Korea Medical Insurance Corporation (KMIC) aged 40 and older who had taken health examination and completed the questionnaire inquiring of health habits and past medical history in 1992 or 1993. The number of cohort members was 698,796, and they were followed up from 1st January, 1994 until 31st December, 2000. The primary sources of the data used in this study were the death benefit record and health examination file of KMIC. In the case that the information about the cause of death was unknown in the death benefit record, it was checked from the death registry of National statistical Office and the inpatient data of KMIC. There were 37439 deaths during the follow-up period. Results: A linear relationship between pulse pressure and mortality from all the causes, cardiovascular disease and cerebrovascular disease, was determined in both genders, the whole population and age groups, in the hypertensive and normotensive ( P＜ 0. 01). Pulse pressure and mortality from all the causes, CVD and CVA increased ( P ＜0.01). Pulse pressure was significantly associated with a relatively high risk of mortality from all the causes, CVD and CVA in the whole population, both genders, all age groups, the hypertensive and normotensive after adjusted to age, gender, body mass index, blood sugar, serum total cholesterol, AST, ALT, urine protein, urine glucose, alcohol drinking and cigarette smoking ( P＜ 0.01). Conclusion: Pulse pressure shows linear relationship with the mortality from all the causes,CVD and CVA. Pulse pressure appears to be a single measure of blood pressure in predicting mortality from all the causes,CVD and CVA, even in the hypertensive and normotensive.