OBJECTIVE The J-curve phenomenon between diastolic blood pressure (DBP) and mortality has been reported repeatedly in treated patients. However, the baseline risk of low DBP has not been fully explored. This study was to examine the relationship between DBP and risk of mortality from all-cause, atherosclerotic vascular diseases (ASCVD), and ischemic heart disease (IHD) using a prospective cohort of general population. METHODS We analyzed 1,234,435 participants of the Korean Cancer Prevention Study cohort (789,255 men, 30-95 years of age) who had a medical evaluation from 1992 to 1995 using Cox proportional hazards models. RESULTS A total of 22.5 million person-years were followed up (mean age 46.6 years, deaths 193,903 cases). The hazard ratios of mortality from all-cause and ASCVD, among those with DBP <60 mmHg compared to 70-79 mmHg were 1.23 (95% confidence interval [CI], 1.16-1.30) and 1.37 (95% CI, 1.20-1.57), respectively, after adjustment for multivariable including systolic blood pressure. Increased risks of all-cause death in the lowest DBP category group were maintained in men or women, 30-59 or ≥60 years of age, smoker or non-smoker and diabetes mellitus (DM) or non-DM subgroups. The risk in DBP 60-69 mmHg groups increased in several subgroups. However, the risk for ASCVD death in 30-59 years and DM group, and risk for IHD death in most subgroups except for elderly (≥60 years) decreased. CONCLUSIONS A J-curve relationship between low DBP and all-cause death was found consistently. The baseline risk in the general population may be considered for risk assessment, particularly in case of interventions that lower DBP below 60 mmHg.