The purpose of this study was to test pathogenetic models for the "unprovoked" ankle inversion injuries seen in functional ankle unstable subjects. The consequence of spatial mal-alignment of the ankle/foot complex on the risk of producing an ankle inversion torque at heel-strike and during swing-phase follow through was analyzed in cadaver simulations. Heel-strike was simulated using a 5 degrees of freedom rig in a material testing machine. A set-up capable of accelerating lower limb specimens towards a support surface simulated swing-phase follow through. Joint excursions were monitored with flexible wire goniometers. The unloaded ankle/foot complex was placed in increasing positions of talar and subtalar joint excursions. The consequences of these settings on the behavior of the ankle/foot complex at heel-strike and when the lateral part of the foot "caught" the ground during swing-phase follow through were monitored. An inversion torque at heel-strike was first seen when the unloaded foot was set in positions exceeding 30 degrees of inversion combined with full plantar flexion and 10 degrees of internal tibial rotation. A collision between the lateral border of a 20 degrees inverted, but otherwise neutral ankle/foot complex and the ground surface during swing-phase follow through forced the foot into the full limit of inversion, plantar flexion and internal tibial rotation measurable in this set-up. Clinical consequence: The study showed that the foot/ankle complex exhibits a high degree of intrinsic stability at heel-strike. The foot will thus stabilize itself and move into normal eversion at the beginning of the stance-phase even though it is set to the ground in a substantial degree of mal-alignment. In contrast, the swing-phase collision model provides a link that can connect the small deficits in inversion angle awareness measured in chronic functional ankle unstable subjects with an increased risk in this group of sustaining ankle inversion injuries.