Background. Lateral ankle instability is a common problem, but the precise role of the lateral ankle structures has not been accurately investigated. This study aimed to accurately investigate lateral ankle complex stability for the first time using a novel robotic testing platform. Method. A six degrees of freedom robot manipulator and a universal force/torque sensor were used to test 10 foot and ankle specimens. The system automatically defined the path of unloaded plantar/dorsi flexion. At four flexion angles: 20° dorsiflexion, neutral flexion, 20° and 40° of plantarflexion; anterior-posterior (90N), internal-external (5Nm) and inversion-eversion (8Nm) laxity were tested. The motion of the intact ankle was recorded first and then replayed following transection of the lateral retinaculum, Anterior Talofibular Ligament (ATFL) and Calcaneofibular Ligament (CFL). The decrease in force/torque reflected the contribution of the structure to restraining laxity. Data were analysed using repeated measures of variance and paired t-tests. Results. The ATFL was the primary restraint to anterior drawer (P< 0.01) and the CFL the primary restraint to inversion throughout range (P< 0.04), but with increased plantarflexion the ATFL's contribution increased. The ATFL had a significant role in resisting tibial external rotation, particularly at higher levels of plantarflexion, contributing 63% at 40° (P< 0.01). The CFL provided the greatest resistance to external tibial rotation, 22% at 40° plantarflexion (P< 0.01). The extensor retinaculum and skin did not offer significant restraint in any direction tested. Conclusion. This study shows accurately for the first time the significant role the ATFL and CFL have in rotational ankle stability. This significant loss in rotational stability may have implications in the aetiology of osteophyte formation and early degenerative changes in patients with chronic ankle instability. This is the first time the role of the lateral ankle complex has been quantified using a robotic testing platform