Paradoxically greater survival for persons aged 85 years and older with higher blood pressures has been reported in a Finnish population study (Br Med J 1988;296:887-9). In a previous report, the authors demonstrated improved 10-year survival with increasing diastolic blood pressure in men (but not in women) aged 75 years and older in the Rancho Bernardo Chronic Disease Study (Br Med J 1989;298:1356-7). However, few of the covariates which could potentially explain this effect were obtained at the visit used in that analysis. In an effort to confirm these reports of paradoxical survival and to explore possible reasons for them, the authors analyzed all-cause and cardiovascular mortality in 795 men and women aged 75-96 years (mean, 80.6), evaluated in 1984-1987 and followed prospectively for an average of 3 years after that comprehensive examination. Of 63 deaths, 48 (76%) were in men; 43 (68%) of all deaths were cardiovascular. Kaplan-Meier survival analyses showed a significant trend for improved survival with increasing diastolic pressure in men aged 80 years and older versus all-cause mortality (chi 2 p less than or equal to 0.01), and cardiovascular mortality (chi 2 p less than or equal to 0.00). These trends were not evident in men aged less than 80 years or in women in either age group. Results were not explained by differences in the use of antihypertensive medication, pulse pressure, history of hypertension, history of coronary heart disease, isolated systolic hypertension, interval change in diastolic pressure (over an average of 12 years), or by cholesterol, triglycerides, fasting plasma glucose, smoking, or body mass index. Thus, the paradoxical relation of improved all-cause and cardiovascular survival in men aged 80 years or older with higher diastolic pressure is not explained by a wide range of biologic and historical factors.