Despite the inverse association between physical activity (PA) and incident hypertension, a comprehensive assessment of the quantitative dose–response association between PA and hypertension has not been reported. We performed a meta-analysis, including dose–response analysis, to quantitatively evaluate this association. We searched PubMed and Embase databases for articles published up to November 1, 2016. Random effects generalized least squares regression models were used to assess the quantitative association between PA and hypertension risk across studies. Restricted cubic splines were used to model the dose–response association. We identified 22 articles (29 studies) investigating the risk of hypertension with leisure-time PA or total PA, including 330 222 individuals and 67 698 incident cases of hypertension. The risk of hypertension was reduced by 6% (relative risk, 0.94; 95% confidence interval, 0.92–0.96) with each 10 metabolic equivalent of task h/wk increment of leisure-time PA. We found no evidence of a nonlinear dose–response association of PA and hypertension (Pnonlinearity=0.094 for leisure-time PA and 0.771 for total PA). With the linear cubic spline model, when compared with inactive individuals, for those who met the guidelines recommended minimum level of moderate PA (10 metabolic equivalent of task h/wk), the risk of hypertension was reduced by 6% (relative risk, 0.94; 95% confidence interval, 0.92–0.97). This meta-analysis suggests that additional benefits for hypertension prevention occur as the amount of PA increases.