Balance is an essential skill for dancers, it helps reduce the risk of injury, and is related to quality of performance. This systematic review aims to investigate the effects of training protocols on the balance of dancers when compared to control groups. Interventional studies, published until January 2023, assessing balance in all levels of ballet, modern, and contemporary dancers were identified in the PubMed, Cochrane, Lilacs, Scielo, Embase, and SPORTDiscus databases. Publications in English, Spanish and Portuguese were considered. In two selection phases, using a standard protocol, two reviewers independently selected and extracted the data. The Downs & Black checklist was used to assess risk of bias. Of the nine intervention studies included, most were classified as either “fair” or “poor” quality. Three-hundred and seventy-two dancers aged 9.6 ± 0.5 to 25.7 ± 3.7 years were analysed. The majority was female (94%), with the years of experience of ballet (35%) or ballet and modern (49%) training ranging from 4.6 ± 1.4 to 12.0 ± 7.5 years. Numerous training protocols of varying durations were found. However, neuromuscular training proved the most effective at improving balance. Most of the studies are classified poor quality. Future studies should consider interventions lasting at least four weeks, two to three times a week, sessions between 30 and 60 min. More high-quality studies, randomized control trials using validated balance tests, as well as new validation studies for dance-specific balance tests are needed. •Different training protocols were found, including physical practice and the use of accessories.•Neuromuscular-based training is an effective strategy to improve dancers' balance.•Interventions should last at least 4 weeks, with 2–3 weekly sessions between 30 to 60 min.•Most of the included studies are classified poor quality.•High-quality studies, RCT using validated balance tests, new validation studies for dance-specific balance tests are needed.