Abstract: Objective To examine the predictive value of cumulative mean arterial pressure（cumMAP）on all-cause death（ACD）. Methods A total of 56 039 subjects who had participated 2006-2007,2008-2009 and 2010-2011 health examinations and had sufficient data were included in the study and followed up for an average of 4.92 years.The information of death in all subjects was collected annually through the Kailuan Social Security Information System. cumMAP was calculated using following formula:cumMAP=（MAP₁+MAP₂）/2×time1-2+（MAP₂+MAP₃）/2×time2-3,where MAP₁,MAP₂,MAP₃ indicated MAP at examination 1,2,3,and time_1-2,time_2-3represented participant-specific time interval between consecutive examinations. Subjects were divided into four groups based on the quartiles of cumMAP. The incidence of all-cause death among four groups were compared and the correlation between cumMAP and all-cause death was analyzed by multivariate Cox proportional hazard regression.Results During a median follow-up of 4.92 years,1805cases all-cause death occurred,accounted for 3.2%. The all-cause mortality was 1.1%,2.3%,3.8%,5.6%in the first,second,third and fourth quartile group respectively（P＜0.01）. Cox proportional hazard regression analysis showed that after adjusted for other confounding factors,including baseline systolic blood pressure and arterial pressure,compared to the first quartile,the HR（95% CI）of all-cause death in second,third and fourth quartile were 1.17（0.95-1.43）,1.32（1.08-1.61）and 1.34（1.09-1.65）. Conclusion cumMAP is an independent risk factor for all-cause death.