Objectives: To determine whether self-reported chronic ankle instability (CAI) affects the physical performance of soldiers during army training, in routine army fitness tests and other sensorimotor tests. Design: Prospective cohort comparison. Methods: Soldiers enrolled on the Infantry Commanders Course (ICCourse) of the Israeli Defense Force were tested in weeks one and fourteen of the course, including a 3000 m timed run, an agility test over an X-shaped course, the Y Balance Test (YBT) and the Active Movement Extent Discrimination Assessment (AMEDA) somatosensory test. Soldiers were trained using routine ICCourse physical training throughout the course. Changes in physical performance were analysed using a two-way ANOVA, for soldier groups categorised as having healthy ankles or CAI. Results: Soldiers improved their scores on all 4 tests (3000 m run, X-agility test, YBT and AMEDA) during the course. However, soldiers with CAI did not improve on the X-agility test, whilst those with healthy ankles did. Soldiers with CAI performed worse on the YBT-PM (postero-medial direction) at the start of the course, but this difference was no longer present at 14 weeks. Soldiers with CAI performed better on the AMEDA at the start and 14-week testing points. Conclusions: Self-reported CAI restricts agility performance gains in soldiers during 14-weeks of physical training but has no effect on 3000 m running performance. Soldiers with CAI perform worse in YBT balance tests. Soldiers with CAI have better proprioceptive acuity than those with stable ankles. These findings indicate possible sensorimotor factors for training of soldiers with known CAI.