To assess the effect of a five-week training program using a wobble board on discrimination of discrete movements at the ankle, 40 ankles were tested from 20 subjects in a group of elite athletes. Players in a first grade rugby league squad were randomly allocated to either a wobble board training or no-training group. Pre-tests to assess discrimination of inversion movements made while standing were carried out on both ankles of all subjects using an automated device that enabled a normal share of weightbearing on the tested ankle. From a sequence of 160 inversion movements, a discrimination index was determined. This measure represents the relative difference in angular extent needed for an inversion movement to be differentiated 75 per cent of the time from the standard movement of 12.5 degrees below horizontal. Subjects were re-tested after either undertaking a five-week wobble board training program, or after having had no training. Discrimination of movements into inversion improved from the pre-test (Weber fraction 8.8 per cent) to the post training test (6.4 per cent) in the wobble board trained group. This discrimination improvement was significantly greater than the change in the untrained group (pre-test: 8.4 per cent and post-test: 8.1 per cent). These data demonstrate that wobble board training can improve discrimination of discrete ankle inversion movements, and support the Weber fraction as a measure of movement sensitivity.