Methods: Searches in five electronic databases were conducted to retrieve studies published from 1980 to 2013. Eligible studies were RCTs consisting of structured exercise training or PA advice versus no intervention in patients with type 2 diabetes. We used random effect models to derive weighted mean differences (WMDs) of exercises on absolute changes in systolic BP (SBP) and diastolic BP (DBP). Results: A total of 30 RCTs of structured training (2,217 patients) and 21 of PA advice (7,323 patients) were included. Data were extracted independently in duplicate. Structured exercise was associated with reductions in SBP (WMD −4.22 mmHg; 95 % confidence interval [CI] −5.89 to −2.56) and DBP (WMD −2.07 mmHg; 95 % CI −3.03 to −1.11) versus controls