Left ventricular hypertrophy (LVH) is an independent cardiovascular risk factor; however, the key strategy necessary for LVH regression in hypertensive patients is not clear. A meta-analysis was conducted to study the effect of blood pressure reduction on LVH regression. We explored the relationship between different degrees of systolic blood pressure (SBP)/diastolic blood pressure (DBP) reduction and LVH regression. A total of 17 randomized controlled trials comprising 2196 hypertensive patients (mean age, 56.3 years; 64.1% were men) were identified. Different degrees of SBP and DBP reductions were significantly associated with LVH regression: SBP reduction ≥20 mm Hg (SBPM20) (weighted mean difference (WMD): 14.35 g m(-2); 95% confidence interval (CI): 10.44, 18.26; P<0.0001); SBP reduction <20 mm Hg (SBPL20 group) (WMD: 14.82 g m(-2); 95% CI: 9.83, 19.8(2); P<0.0001); DBP reduction ≥10 mm Hg (DBPM10 group) (WMD: 15.17 g m(-2); 95% CI: 11.86, 18.48; P<0.0001); and DBP reduction <10 mm Hg (DBPL10 group) (WMD: 11.76 g m(-2); 95% CI: 3.75, 19.76; P=0.004). Significant regression of LVH was found in the DBPM10 group compared with the SBPM20, SBPL20 and DBPL10 groups (P<0.0001). The most significant decrease in LVH was seen in patients with a mean age over 60 years in the DBPM10 group. Moreover, the renin-angiotensin system inhibitor was found to be the most effective antihypertensive drug for LVH regression. This meta-analysis result indicates that proper DBP reduction plays an important role in the regression of echocardiographic LVH in hypertensive patients.