The body roundness index (BRI) has good predictive ability for both body fat and visceral adipose tissue. Longitudinal BRI trajectories can reveal the potential dynamic patterns of change over time. This prospective study assessed potential associations between BRI trajectories and incident cardiovascular disease (CVD) in rural regions of Northeast China. In total, 13,209 participants (mean age: 49.0 ± 10.3 years, 6,856 [51.9%] male) were enrolled with three repeated times of BRI measurements at baseline (2004-2006), 2008, and 2010, and followed up until 2017 in this prospective study. Using latent mixture model, the BRI trajectories were determined based on the data from baseline, 2008 and 2010. Composite CVD events (myocardial infarction, stroke, and CVD death combined) was the primary endpoint. Cox proportional-hazards models were used to analyze the longitudinal associations between BRI trajectories and incident CVD. Three distinct BRI trajectories were identified: high-stable (n = 538), moderate-stable (n = 1,542), and low-stable (n = 11,129). In total, 1,382 CVD events were recorded during follow-up. After adjustment for confounders, the moderate-stable and high-stable BRI groups had a higher CVD risk than did the low-stable BRI group, and the HR (95%CI) were 1.346 (1.154, 1.571) and 1.751 (1.398, 2.194), respectively. Similar associations were observed between the trajectories of BRI and the risk of stroke and CVD death. The high-stable group was also significantly and independently associated with CVD, myocardial infarction, stroke, and CVD death in participants aged <50 years. BRI trajectory was positively associated with incident CVD, providing a novel possibility for the primary prevention of CVD in rural regions of China.