Objectives: To systematically review and synthesize the effects of soft braces on pain and on self-reported and performance-based physical function in patients with knee osteoarthritis. Data Sources: The following electronic databases were searched from inception to April 20, 2016: The Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, EMBASE, CINAHL, SPORTDiscus, Web of Science, and PEDro. Study Selection: Randomized controlled trials (RCTs) and nonrandomized controlled trials (non-RCTs), such as controlled clinical trials, crossover studies, and case-control studies, were included. Two reviewers independently screened articles and determined inclusion through predefined criteria. Data Extraction: Data related to participant demographics, study design and methods, interventions, and outcomes, including numerical means and SDs, were extracted by 1 reviewer. Methodological quality assessment was independently performed by 2 reviewers. Data Synthesis: Eleven studies were identified, including 6 RCTs and 5 non-RCTs. The methodological quality of included RCTs was low. There was a moderate improvement in pain (standardized mean difference [SMD]=.52; 95% confidence interval [CI], .14-.89; P=.007; 284 participants) in favor of wearing a brace compared with not wearing a brace for the immediate, within-group comparison. There was a moderate improvement in pain (SMD=.61; 95% CI, .33-.89; P<.001; 206 participants) and a small to moderate improvement in self-reported physical function (SMD=.39; 95% CI, .11-.67; P=.006; 206 participants) in favor of patients receiving a soft brace versus standard care for the prolonged effect, between-group comparison. Conclusions: Currently available evidence indicates that soft braces have moderate effects on pain and small to moderate effects on self-reported physical function in knee osteoarthritis. These findings highlight the importance of soft braces as a technique to improve pain and physical function in both the short- and long-term. Additional high-quality studies are warranted to improve confidence in the findings.