BACKGROUND AND OBJECTIVE: Physical activity has benefits for the cardiovascular system, however, what levels and types of activity provide optimal cardiovascular health is unclear. We aimed to determine the level of physical activity that has the most benefits against cardiovascular diseases (CVD). METHODS: PubMed, Scopus, and Web of Science were searched for prospective cohort studies on leisure-time (LTPA) or occupational physical activity (OPA) as the exposure and major types of CVD (total CVD, coronary heart disease [CHD], stroke, and atrial fibrillation [AF]) incidence as the outcome. Risk of bias of studies was evaluated using the ROBINS-I tool. Summary hazard ratios (HR) were calculated using random-effects pairwise model. RESULTS: A total of 103 studies were included in the analysis. The highest versus the lowest LTPA was associated with a lower risk of overall CVD (HR = 0.81; 95% CI: 0.77-0.86), CHD (HR = 0.83; 0.79-0.88), and stroke (HR = 0.83; 0.79-0.88), but not AF (HR = 0.98; 0.92-1.05). Linear dose-response analyses showed a 10%, 12%, 9%, and 8% risk reduction in CVD, CHD, stroke, and AF incidence, respectively, for every 20 MET-hours/week increase in LTPA. In nonlinear dose-response analyses, there were inverse associations up to 20 MET-hours/week with 19% and 20% reduction in CVD and CHD risk, and up to 25 MET-hours/week with 22% reduction in stroke, with no further risk reduction at higher LTPA levels. For AF, there was a U-shaped nonlinear association with the maximum 8% risk reduction at 10 MET-hours/week of LTPA. Higher levels of OPA were not associated with risk of CVD, CHD, stroke, or AF. CONCLUSIONS: Overall, results showed an inverse dose-response relationship between LTPA and risk of CVD, CHD, stroke, and AF. Running was the most beneficial LTPA but the risk was similar among various LTPA intensities. OPA showed no benefits in total or any type of CVD.