OBJECTIVES: To examine associations between objective measures of activity level and mortality risk in older men. DESIGN: Prospective cohort study. SETTING: Six U.S. sites. PARTICIPANTS: Men aged 71 and older followed an average of 4.5 years (N = 2,918). MEASUREMENTS: Time awake spent in sedentary behavior (metabolic equivalent (MET) level ≤1.50), light activity (MET level 1.51-2.99), and at least moderate activity (MET level ≥3.00) measured using an activity monitor worn for 5 days or longer and expressed as quartiles. Deaths were confirmed with death certificates; cause of death was adjudicated by review of certificates and records. RESULTS: During follow-up, 409 (14%) men died. After multivariable adjustment, comparing Q4 with Q1, more time spent in sedentary behavior (Q4 vs Q1, hazard ratio (HR) = 1.51, 95% confidence interval (CI) = 1.10-2.08), less time spent in light activity (Q1 vs Q4, HR = 1.54, 95% CI = 1.06-2.24), and less time spent in at least moderate activity (Q1 vs Q4, HR = 1.56, 95% CI = 1.09-2.25) were similarly associated with greater mortality risk primarily due to higher risks of cardiovascular and noncardiovascular, noncancer death. The association between time spent in sedentary behavior and mortality varied according to time spent at higher activity level. More time spent in sedentary behavior was associated with greater risk of death in men spending 1.2 (median) h/d or more in at least moderate activity (Q4 vs Q1, HR = 2.09, 95% CI = 1.26-3.49) but not in those spending less time (Q4 vs Q1, HR = 1.02, 95% CI = 0.62-1.66) (P = .005 for interaction). CONCLUSION: In older men exceeding current guidelines on physical activity, more time spent in sedentary behavior is associated with greater mortality risk.