While previous studies have investigated the effect of shoe–surface interaction on injury risk, few studies have examined the effect of rotational stiffness of the shoe. The hypothesis of the current study was that ankles externally rotated to failure in shoes with low rotational stiffness would allow more talus eversion than those in shoes with a higher rotational stiffness, resulting in less severe injury. Twelve (six pairs) cadaver lower extremities were externally rotated to gross failure while positioned in 20 deg of pre-eversion and 20 deg of predorsiflexion by fixing the distal end of the foot, axially loading the proximal tibia, and internally rotating the tibia. One ankle in each pair was constrained by an American football shoe with a stiff upper, while the other was constrained by an American football shoe with a flexible upper. Experimental bone motions were input into specimen-specific computational models to examine levels of ligament elongation to help understand mechanisms of ankle joint failure. Ankles in flexible shoes allowed 6.7±2.4 deg of talus eversion during rotation, significantly greater than the 1.7±1.0 deg for ankles in stiff shoes (p = 0.01). The significantly greater eversion in flexible shoes was potentially due to a more natural response of the ankle during rotation, possibly affecting the injuries that were produced. All ankles failed by either medial ankle injury or syndesmotic injury, or a combination of both. Complex (more than one ligament or bone) injuries were noted in 4 of 6 ankles in stiff shoes and 1 of 6 ankles in flexible shoes. Ligament elongations from the computational model validated the experimental injury data. The current study suggested flexibility (or rotational stiffness) of the shoe may play an important role in both the severity of ankle injuries for athletes.