We explored the bidirectional association between depressive symptoms and measures of carotid atherosclerosis. This study included 1155 participants aged 55 years and older. Depressive symptoms was assessed by 30-item Geriatric Depression Scale (GDS-30) score. Common carotid artery intima-media thickness (CCA-IMT), carotid plaques and carotid stenosis were measured at both common carotid arteries using an edge detection system. In part 1, we explored the risk of CCA-IMT on depressive symptoms. We identified a significantly increased risk of depressive symptoms with the severity of carotid atherosclerosis. ORs (95%CI) of CCA-IMT thickening, carotid plaque and carotid stenosis for depressive symptoms were 1.48(0.72–3.03), 2(1.03–3.85) and 5.29(2.16–12.97) comparing with normal CCA-IMT in adjusting all potential risk factors models. When using carotid atherosclerosis as a continuous variable, the OR for depressive symptoms was 1.32 (95%CI 1.16–1.49) with every 0.1mm elevated in CCA-IMT after adjustment for all potential confounders. In part 2, we explored the risk of depressive symptoms on carotid atherosclerosis. We found elevated depressive symptoms significantly increased the risk of carotid atherosclerosis in multi-factor adjusted models [OR (95%CI): 1.65(1.10–2.47)]. When using depressive symptoms as a continuous variable, the ORs for carotid artery atherosclerosis were 1.32 (95%CI 1.16–1.49) with every 1 GDS-30 score elevated after adjustment for all convenient risk factors. The current study demonstrated the bidirectional links of carotid atherosclerosis measures with depressive symptoms. More cohort study and clinical trial focusing on the issue need to be explored in the future. © 2019 Elsevier B.V.