Several studies have indicated that impaired postural stability is associated with an increased risk of falling in older adults. Traditional exercise programs consisting of stretching and full body movements have typically failed to show an effect on stability. Therefore, it has been recommended that training programs be customized to target the specific physiological systems involved in balance control, specifically the visual, vestibular, somatosensory, motor, and musculoskeletal systems. To determine the effectiveness of such a program, 13 older adults aged 79–90y (84.2y ± 3.8; mean ± SD) living in a retirement community participated in a 10-wk (3d/wk, 50 min/d) balance training program. Training was performed standing on 2ö- and 4ö-thick upholstery foam with the eyes open (EO) or eyes closed (EC), stepping over foam obstacles, walking on 12' foam pathways, standing from a chair, climbing four stairs, and walking around cones. To determine postural sway, participants completed the modified CTSIB on the NeuroCom Balance Master System. Tests consisted of three 10 s trials on four conditions performed with bare feet: EO-Firm surface, EC-Firm surface, EO-Foam (5ö thick) surface, EC-Foam surface. A 2 × 2 × 2 (Time × Surface × Vision) randomized block design revealed a three-way interaction (F = 12.66, p < 0.005). Of more interest were significant findings for Time × Surface (F = 9.85, p < 0.01) and non-significant findings of Time × Vision (F = 0.88, p = 0.37). Results suggest the balance program was successful in reducing sway while standing on different surfaces, but did not affect sway during the visual conditions. Other significant findings include Surface × Vision (F = 33.23, p < 0.001) and Time, Surface, and Vision main effects (F = 14.67, F = 54.15, F = 147.87, respectively; p < 0.001). Further studies are needed to evaluate the effects of similar training protocols on balance, functional performance, and the incidence of falls.