Some studies have shown that a higher ideal cardiovascular health score (CVHS) predicts a lower incidence of arterial stiffness. Few studies have used multiple measurements of CVHS to examine the impact of CVHS on arterial stiffness. The current study aimed to identify the long-term patterns in CVHS trajectory and to explore the association between CVHS trajectory and arterial stiffness. The study cohort consisted of 18,854 participants from the Kailuan Study who were followed up for five physical examinations over 8.10 years. Five discrete CVHS trajectories were identified among the participants: low-stable (8.10%), low-moderate (6.84%), moderate-low (23.46%), moderate-stable (39.83%), and elevated-stable (21.77%). After adjustment for confounding factors, generalized linear model analysis showed that CVHS trajectory group correlated negatively with brachial–ankle pulse wave velocity (baPWV). Compared with the low-stable group, the low-moderate group, moderate-low group, moderate-stable group, and elevated-stable group had B values of −41.81, −24.11, −86.79, and −169.54, respectively. We also used logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for CVHS trajectory groups and arterial stiffness. In fully adjusted models, ORs were 0.76 (95% CI: 0.62–0.94) for the low-moderate group, 0.80 (95% CI: 0.67–0.97) for the moderate-low group, 0.51 (95% CI: 0.42–0.62) for the moderate-stable group, and 0.23 (95% CI: 0.18–0.29) for the elevated-stable group compared with the low-stable group. The results were consistent across a number of sensitivity analyses. In conclusion, the higher long-term attainment and the improvement of CVHS were negatively associated with baPWV and could reduce the risk of arterial stiffness. Our study emphasizes the importance of optimizing CVH throughout life to prevent the incidence of arterial stiffness. © 2019, The Japanese Society of Hypertension.