Physical exercise has been recognized as an important strategy in the promotion of healthy aging. Positive effects on older adults’ motor ability are brought about by engaging their motor skills and promoting sensorimotor learning and cortical plasticity. These processes could be increased with the use of immersive virtual reality (IVR) technology, since the multisensory stimulation is greater. The aim of this study was to explore the usability and balance effects of an IVR exercise program in community-dwelling nonagenarian people. A sample of 12 women were allocated to an experimental group (EG n = 6; 91.67 ± 1.63 years) and a control group (CG n = 6; 90.83 ± 2.64 years). For 10 weeks, the EG used a commercial IVR exergame three times a week. All the sample completed the program without adverse effects (without Simulator Sickness Questionnaire symptoms). Post-gaming usability was good (System Usability Scale 78.33). The EG improved some balance parameters significantly (Tinetti test: balance (10.97 %; Sig = 0.017), gait (9.23%; Sig = 0.047) and total score (10.20%; Sig = 0.014) and maintained total TUG test times (−0.45%)). There were significant differences between groups (Tinetti test: balance (Sig = 0.004) and total score (Sig = 0.0032)). We successfully demonstrated that IVR training is feasible and is an effective and personalized method to enhance balance and to reduce the risk of falls in community-dwelling nonagenarian women.