The effectiveness of regular exercise in decreasing both systolic and diastolic blood pressure (BP) is well-established. Our purpose was to assess the impact of both subject and physical activity program characteristics on ambulatory BP through a meta-analysis of the existing literature. Three databases (PubMed, Embase, Web of Science) were searched using relevant terms and strategies. From 637 identified records, 37 studies met inclusion criteria: details about training intervention and participant characteristics, pre- and post-training ambulatory BP measurements, at home (HBPM) or during 24-h (ABPM). The weighted mean difference was for 24-h systolic/diastolic ABPM (n = 847 participants): -4.06/-2.77 mmHg (95%CI: -5.19 to -2.93/-3.58 to -1.97; P < 0.001), for daytime (ABPM or HBPM, n = 983): -3.78/-2.73 mmHg (95%CI: -5.09 to -2.47/-3.57 to -1.89; P < 0.001) and nighttime ABPM periods (n = 796): -2.35/-1.70 mmHg (95%CI: -3.26 to -1.44/-2.45 to -0.95; P < 0.001). Characteristics significantly influencing BP improvement were: an initial office BP >/=130/85 mmHg and diet-induced weight-loss. We found no differences according to sex, age, or training characteristics (intensity, number of sessions, training duration). Antihypertensive effects of aerobic training assessed by ambulatory BP measurements appear significant and clinically relevant for both daytime and nighttime periods, in particular for participants with an office BP >/=130/85 mmHg.