The effects of two training programs on movement discrimination ability, at the ankle and knee, were assessed from the left and right lower limbs of forty-four football players. All players in three Under 18 Victorian Football League (VFL) squads were allocated to either wobble board training, jump landing training, or no-training conditions. Pre-tests to assess discrimination of extent for active movements made while standing were carried out on both ankles and knees of all subjects, using an automated device to accurately set the different movement stop points. Five distances were used, between 10.5 degrees and 14.5 degrees from horizontal for ankle inversion, and between 30.3 degrees and 31.7 degrees from vertical for knee flexion. From a series of 50 inversion movements and 50 knee flexion movements, matrices of absolute judgement by actual movement extent were produced. Non-parametric signal detection analysis was applied to the discrimination score. All subjects were retested after eight weeks. Improvement in discrimination of ankle movements into inversion from pre-test (0.65) to post test (0.70) for the wobble board trained group was significantly larger than the change in the jump-landing trained and the untrained groups (Jump Landing: Pretest: 0.64 to Post-test: 0.64 and Control; Pretest: 0.63 to Post-test: 0.64). Discrimination of knee flexion movements improved significantly from pre-test to post-test in all three groups. These data demonstrate that wobble board training can improve discrimination of discrete ankle inversion movements, an effect interpreted as enabling greater accuracy in the making of inversion movements in foot preparation prior to ground contact.