Introduction Commonly used stress tests require radiography. The respective reliability and sensitivity is still under debate. We developed a non-radiographic ankle arthrometer to objectively assess mechanical ankle stability. The aim of the present study was to validate the ankle arthrometer in a cadaver and in vivo set up.

Methods Cadaver study Nine cadaver specimens were tested under different anterior drawer load and ankle angle positions against standardised radiography. Lateral ankle ligaments were sequentially cut during testing procedure. Stiffness was calculated and the diagnostic validity of the ankle arthrometer was analysed.

In vivo study 23 subjects participated in the study. An orthopaedic surgeon first performed a manual anterior talar drawer test to classify the subjects’ ankles as stable or unstable. Subjects then underwent ankle arthrometer testing up to 200 N anterior drawer load twice and the stiffness were calculated.

Results Cadaver study Stable and unstable ankles were significantly differentiated by stiffness analyses (p=0.01) with the ankle arthrometer adjusted to 0 degrees of plantarflexion and 50 N anterior drawer load application.

Results from the ankle arthrometer radiographs and the standardised stress testing device correlated significantly (r=0.73–0.94; p=0.00–0.03). Stiffness analyses discriminate stable and unstable ankles with a sensitivity of 91.7% and specificity of 62.5%.

In vivo study Using the manual anterior drawer test, 16 ankles were classified as stable and 7 as unstable. Stiffness group analysis differentiated stable from unstable ankles significantly (p=0.00 and p=0.01). Test-retest demonstrated an accurate reliability (intraclass correlation coefficient=0.80).

Conclusions Validation and reliability of the non-radiographic ankle arthrometer was demonstrated in a cadaver and in vivo set up. The ankle arthrometer stiffness analyses may be relevant to screen athletes at risk for lateral ankle instability, for follow-up studies, and to implement preventive strategies.