We wanted to use biomechanical testing in a cadaveric model to compare the Brostrom repair, the Watson-Jones reconstruction, and a new anatomic reconstruction method. Eight specimens were held in a specially designed testing apparatus in which the ankle position (dorsiflexion-plantar flexion and supination-pronation) could be varied in a controlled manner. Testing was done with intact ligaments and was repeated after sectioning of the anterior talofibular ligament and the calcaneofibular ligament and after a Brostrom repair, a Watson-Jones reconstruction, and a new anatomic reconstruction were performed. An antedor drawer test was performed using an anterior translating force of 10 to 50 N, and a talar tilt test was performed using a supination torque of 1.1 to 3.4 N-m. The forces in the anterior talofibular ligament and calcaneofibular ligament were measured with buckle transducers, and tibiotalar motion and total ankle joint motion were measured with an instrumented spatial linkage. The increase in ankle joint laxity observed after sectioning of both the anterior talofibular and catcaneofibular ligaments was significantly reduced by the three reconstructive techniques, although not always to the level of the intact ankle. Joint motion was restricted after the Watson-Jones procedure compared with that in the intact ankle. Unlike the Watson-Jones procedure, theligament or graff force patterns observed during loading after the Brostrom repair and the new anatomic technique resembled those observed in the intact ankle.