Background Strength testing of the ankle musculature using the isokinetic dynamometer is often undertaken by sports clinicians and is of great importance for screening, rehabilitation and injury prevention. Injuries such as ankle sprains may be preventable if the risk factors can be addressed; however, the success of such screening procedures depends on the accuracy and reproducibility of the methods used.

Objective To determine the test-retest reliability of three setup positions during isokinetic ankle inversion-eversion exercise, and to investigate the effect of setup position on peak torque and total work.

Design Randomised test-retest study.

Setting University biomechanics laboratory. Participants played at a semi-professional level of soccer.

Participants: A convenience sample was used to recruit sixteen healthy male participants. Inclusion criteria included males, aged 18–25 years, who participated in semi-professional soccer and who were right leg dominant.

Risk factor assessment Three maximal concentric ankle inversion-eversion repetitions at 60, 120, 180, 240, 300 and 360° · s⁻¹, during 10° dorsiflexion, neutral dorsiflexion/plantarflexion, and 10° plantarflexion. Set up position and isokinetic speed were the independent variables.

Main outcome measure Peak torque and total work in newton metres in the three setup positions.

Results The results indicated that the 10° plantarflexion position was the most reliable setup, with ICC results ranging from 0.84–0.95 for peak torque and total work, at speeds 60 through 240°· s⁻¹, during ankle inversion and eversion. The results also showed significantly greater (P<.0167) peak torque and total work values for the 10° plantarflexion position.

Conclusion Clinicians should consider adopting this new ankle setup when screening or rehabilitating athletes as the most reliable results were identified. The 10° plantarflexion position also most accurately represented the peak performance of the muscles tested, as peak torque and total work were maximised in this position.