Context: The sixth Joint National Committee (JNC-VI) classification system of blood pressure emphasizes both systolic blood pressure (SBP) and diastolic blood pressure (DBP) for cardiovascular disease risk assessment. Pulse pressure may also be a valuable risk assessment tool. Objective: To compare relationships of SBP, DBP, and pulse pressure, separately and jointly, with cardiovascular disease-related mortality in men. Design and Setting: Data from the Multiple Risk Factor Intervention Trial (MRFIT), which screened men aged 35 to 57 years from 1973 through 1975 at 22 US centers, was used to assess cardiovascular disease-related mortality through 1996. Participants: A total of 342815 men without diabetes or a history of myocardial infarction were divided into 2 groups based on their age at MRFIT screening (35- to 44-year-olds and 45- to 57-year olds). Participant blood pressure levels were classified into a JNC-VI blood pressure category based on SBP and DBP (optimal, normal but not optimal, high normal, stage 1 hypertension, stage 2-3 hypertension), and pulse pressure was calculated. Main Outcome Measure Cardiovascular disease-related mortality. Results: There were 25721 cardiovascular disease-related deaths. Levels of SBP and DBP were more strongly related to cardiovascular disease than pulse pressure. Relationships of SBP, DBP, and pulse pressure to cardiovascular disease-related mortality varied within JNC-VI category. Concordant elevations of SBP and DBP were associated with a greater risk of cardiovascular disease-related mortality for both age groups of men. Among men aged 45 to 57 years, higher SBP and lower DBP (discordant elevations) also yielded a greater risk of cardiovascular disease-related mortality. Conclusion: In both age groups, cardiovascular disease risk assessment was improved by considering both SBP and DBP, not just SBP, DBP, or pulse pressure separately.