The computerized tomography (CT) data of a hallux valgus patient were selected to establish a finite element model. The standard Scarf osteotomy was simulated based on the model, and the rotation and translation were performed, respectively. The relevant parameters of hallux valgus were compared between the two groups. We completed the cartilage, ligament and other tissues on the bone model to establish a full foot model. We analyzed the troughing, plantar aponeurosis tension, plantar soft tissue, and ground stress and also observed the stability of the broken end by a three-point bending test. The results show that: rotational osteotomy can correct larger IMA, and the contact area after osteotomy is larger. The length of the first metatarsal initially increased and then decreased. Compared with the translational group, rotational Scarf can significantly increase DMAA. Mechanical analysis showed that the cancellous bone in the contact part of the broken end in the translation group had greater stress, which was the troughing performance. Stress distribution of plantar aponeurosis, plantar soft tissue, and the ground showed no significant difference. The three-point bending test showed that the separation of the broken ends of the rotational Scarf osteotomy model was slightly smaller than the translational group. We believe that rotational Scarf osteotomy can correct larger IMA, significantly increase the broken end contact area, and effectively avoid troughing. It also had broken end separation smaller than the traditional translational Scarf osteotomy, which is conducive to postoperative stability and healing. However, rotational Scarf osteotomy also has disadvantages, which can increase DMAA accordingly. For patients with large DMAA, other surgical methods should be considered clinically.