Background: Falls are the leading causes of (non)fatal injuries in older adults. Recent research has developed interventions that aim to improve balance in older adults using virtual reality (VR).Purpose: We aimed to investigate the validity, reliability, safety, feasibility, and efficacy of head mounted display (HMD) systems for assessing and training balance in older adults.Methods: We searched EBSCOhost, Scopus, Web of Science, and PubMed databases until 1 September 2020 to find studies that used HMD systems for assessing or training balance. The methodological quality was assessed using a modified version of Downs and Black. We also appraised the risk of bias using Risk of Bias Assessment tool for Non-randomized Studies (RoBANS).Results: A total of 19 articles (637 participants) were included for review. Despite heterogenous age ranges and clinical conditions across studies, VR HMD systems were valid to assess balance and could be useful for fall prevention and for improving postural control and gait patterns. These systems also have the capacity to differentiate healthy and balance-impaired individuals. During VR versions of traditional balance tests, older adults generally acquire a cautious behavior and take more time to complete the tasks.Conclusion: VR HMD systems can offer ecologically valid scenarios to assess and train functional balance and can be used alone or in addition to other interventions. New norms and protocols should be defined according to participants' age, health status, and severity of their illness when using VR HMD systems for balance assessment and training. For safe and feasible training, attention must be given to display type, VR elements and scenarios, duration of exposure, and system usability. Due to high risk of bias and overall poor quality of the studies, further research is needed on the effectiveness of HMD VR training in older adults.