Background: Depressive symptoms tend to fluctuate over time. Data on the relationship between time-dependent depressive symptoms and the risk of cardiovascular and all-cause mortality among the elderly in China are lacking. Methods and results: A prospective cohort of 1999 subjects aged >= 55 years were enrolled in the Beijing Longitudinal Study of Aging since 1992. Depressive symptoms were assessed at baseline (0 years) and after 2, 5, 8, and 12 years, defined as a score of >= 16 on the 20-item Center for Epidemiological Studies Depression Scale. Mortality status was obtained from the local death registry until December 31st, 2012. Hazard ratio (HR) for all-cause mortality and sub-distribution HR (SHR) for cardiovascular mortality were respectively deduced from time-dependent Cox and competing risk models. During 19,658 person-years of follow-up, 1127 (55.65%) deaths were recorded, of which 483 (23.85%) were attributable to cardiovascular inclinations. Baseline depressive symptoms were neither associated with all-cause mortality (adjusted HR: 1.12, 95% confident interval, CI: 0.94-1.33) nor cardiovascular mortality (adjusted SHR: 1.10, 95% CI: 0.82-1.46) after adjustment of potential cardiac-risk factors. When depressive symptoms were used as time-dependent variable updated from 1992 to 2004, the associations were significant for both all-cause mortality (adjusted HR: 1.48, 95% CI: 1.26-1.73) and cardiovascular mortality (adjusted SHR: 1.40, 95% CI: 1.08-1.82) in the full adjusted model. Conclusions: Time-dependent depressive symptoms increased the risk of all-cause and cardiovascular mortality among the elderly in China. (C) 2018 Japanese College of Cardiology. Published by Elsevier Ltd. All rights reserved.