We compared systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP), and mean arterial pressure (MAP) in predicting the risk of cardiovascular disease (CVD), stratifying results at age 60 years, when DBP decreases while SBP continues to increase. We prospectively followed 11 150 male physicians with no history of CVD or antihypertensive treatment through the 2-year questionnaire, after which follow-up began. Reported blood pressure was averaged from both the baseline and 2-year questionnaires. During a median follow-up of 10.8 years, there were 905 cases of incident CVD. For men aged <60 years (n=8743), those in the highest versus lowest quartiles of average SBP (>/=130 versus <116 mm Hg), DBP (>/=81 versus <73 mm Hg), and MAP (>/=97 versus <88 mm Hg) had relative risks (RRs) of CVD of 2.16, 2.23, and 2.52, respectively. Models with average MAP and PP did not add information compared with models with MAP alone (P>0.05). For men aged >/=60 years (n=2407), those in the highest versus lowest quartiles of average SBP (>/=135 versus <120 mm Hg), PP (>/=55 versus <44 mm Hg), and MAP (>/=99 versus <91 mm Hg) had RRs of CVD of 1.69, 1.83, and 1.43, respectively. The addition of other blood pressure measures did not add information compared with average SBP or PP alone (all P>0.05). These data suggest that average SBP, DBP, and MAP strongly predict CVD among younger men, whereas either average SBP or PP predicts CVD among older men. More research should distinguish whether MAP, highly correlated with SBP and DBP, better predicts CVD.