Aim: This meta-analysis aims at comparing motor improvement after subthalamic (STN) and pallidal (GPi) deep brain stimulation (DBS) for patients with Parkinson's disease (PD). Methods: We searched PubMed through November,2014 for prospective controlled studies comparing STN DBS and Gpi DBS for PD patients. Changes in UPDRS motor score, activities of daily life, verbal fluency score and levodopa equivalent dose were pooled as standardized mean difference between two groups in a meta-analysis model using RevMan 5.3. Results: Nine controlled trials with a total of 497 patients were eligible for this study. The overall effect did not favor either of the two groups in terms of improvement in UPDRS motor score (off medication SMD ¼ 0.11, 95% CI ¼ [0.30, 0.09] and on medication SMD ¼ 0.04, 95% CI ¼ [0.15, 0.23]), activities of daily life (SMD ¼ 0.10, 95% CI ¼ [0.31, 0.11]), semantic verbal fluency (SMD ¼ 0.04, 95% CI ¼ [0.25, 0.16]) and phonemic verbal fluency (SMD ¼ 0.15, 95% CI ¼ [0.35, 0.06]). The levodopa equivalent dose was less in patients undergoing STN DBS than GPi DBS (SMD ¼ 0.29, 95% CI ¼ [0.48, 0.10]). Conclusion: STN DBS allows more reduction in medication than GPi DBS. Subthalmic and Pallidal DBS achieved the same motor improvement in PD patients, so we recommend that choosing surgical target in PD patients should be based on other non-motor outcomes.