Tai Chi has frequently been used as a preventive measure against falling in at-risk populations. However, studies have yielded contradictory results, and literature reviews have considered only a small number of trials and have not addressed some key aspects, such as sources of heterogeneity and publication bias. This study includes 13 controlled trials published before June 2015 that analyzed the effectiveness of Tai Chi in fall prevention in populations of frail and at-risk adults. The effect measure used in this meta-analysis was absolute risk reduction (ARR) with a 95% confidence interval (CI). According to our findings, practice of Tai Chi significantly prevents the risk of falling (ARR, [Formula: see text]; 95% CI: [Formula: see text], [Formula: see text]). The heterogeneity of results across the trials was low, with a reduced risk of publication bias, and no significant effect differences were observed between studies comparing Tai Chi with other interventions or non-treatment. We therefore conclude that Tai Chi is more effective than other measures, or no intervention, for fall prevention in at-risk populations. Further research is warranted to analyze the consequences of falls and to study the episodes rather than the cases of falls.