In this study, ankle joint stability was evaluated using multibody computational ankle joint model to assess two new anatomic reconstruction and three popular non-anatomic (Watson-Jones, Evans, and Chrisman-Snook) reconstruction techniques. An LAS injury model, three popular non-anatomic reconstruction models (Watson-Jones, Evans, and Chrisman-Snook), and two common types of anatomic reconstruction models were developed based on the previously validated ankle joint model. The stability of ankle in both talocrural and subtalar joint were evaluated under anterior drawer test, inversion test, internal rotational test, and the combined loading test to mimic ankle sprain mechanism. Overall results show that the two anatomic reconstruction techniques were superior to the non-anatomic reconstruction techniques in stabilizing both talocrural and subtalar joints. The loss of subtalar joint motion, which mainly observed in Watson-Jones and Chrisman-Snook techniques for both single and combined loading conditions, was not shown in the anatomical reconstructions. Among non-anatomic reconstructions, Evans technique was beneficial for subtalar joint as it does not restrict subtalar motion, though Evans technique was insufficient for restoring talocrural joint inversion. The anatomical reconstruction techniques best recovered ankle stability to the level seen in the intact ankle joint.