Vestibular rehabilitation is effective and safe in patients with instability. However, there is insufficient evidence for distinguishing between efficacies of different dosage of therapies. Therefore, the aim of the present study was to verify whether there were differences between two computerised dynamic posturography (CDP) therapies of different numbers of sessions, in order to establish the optimal strategy. We conducted a prospective, comparative study of two different dosage of CDP therapy (a 5-session group and another of 10-session group) in patients with instability due to chronic unilateral peripheral vestibular disorder. We used balanced block randomisation to include 13 patients in each group. Improvement was assessed using the Dizziness Handicap Inventory and the CDP with the sensorial organisation test (SOT) and limits of stability (LOS). We found a statistically significant improvement in both groups in composite score, visual and vestibular input (SOT); and in reaction time, distance and directional control (LOS). If we compare the groups regarding these improvements, we found that 10-session group showed a greater benefit in distance covered and directional control of LOS. Since significant improvement is obtained with only five sessions, we believe this to be the optimal number of treatment sessions for most patients with chronic unilateral peripheral vestibular disorder. Nevertheless, those patients with more reduced limits of stability, and consequently greater likelihood of falling as a result of their diminished base of support, are candidates for rehabilitation protocols with a greater number of sessions.