Compared with resistance training, information concerning the progressive configuration of balance training (BT) is rare and lacks scientific validation. Therefore, a study was designed to determine participants' ability to perform balance exercises with increasing level of difficulty. The task required the participants (N = 20) to stand as stable as possible on a computerized balance platform. The experiment was performed on 3 testing days using different stance and sensory conditions. On each day, bipedal, step, tandem, and monopedal stands were performed 3 times while sensory conditions changed from firm ground, eyes opened (day 1) over foam ground, eyes opened (day 2) to firm ground, eyes closed (day 3). The results showed that total center of pressure displacements significantly increased when the use of sensory information (comparison between testing days: all p < 0.001) or when the base of support (comparison within testing days: all p < 0.001) was gradually reduced. Based on the observed pattern of increased postural sway across all testing conditions and the levels of trial variability, exercises were categorized into several stages of training. Findings indicate that balance performance decreased in response to an increased level of task difficulty introduced by narrowing the base of support and by limiting the use of sensory information. Practitioners can use the derived exercise ranking to select exercises for BT appropriate to the level of participants' balance ability and to implement progression in balance training.