Background: The preoperative identification of patients who might need constrained condylar knee (CCK) prosthesis in total knee arthroplasty (TKA), is essential to ensure the availability of equipment and to address the patients’ expectations accurately. Objectives: In this study, we aimed at investigating if the preoperative features of the patients can provide this data. Methods: A total of 30 patients who underwent primary TKA for severe osteoarthritic genu varum deformity (varus angle ≥20º) were evaluated in this retrospective study. Prosthesis selection was based on preoperative and intraoperative information. Demographic data, preoperative correctability of the deformity, and intraoperative information, including the reduction osteotomy, soft-tissue release, and pie-crust technique, were retrospectively collected. Soft-tissue release was performed in a sequential manner in 3 steps. Results: The study population included 4 males and 26 females with a mean±SD age of 64.6±8.7 years. A CCK prosthesis was used in 11 (36.7%) cases. A significant association was found between the preoperative correctability and the type of prosthesis. In other words, all CCK prostheses were used in patients who were preoperatively non-correctable (P<0.001). Also, the step of release was significantly associated with the type of prosthesis, and CCK prosthesis was used in all patients with step 3 release (P<0.001). Preoperative correctability was significantly related to the step of release, as well. It means that all deformities with step 3 release were preoperatively non-correctable (P=0.008). Conclusion: The preoperative clinical evaluation of correctability could be used in the identification of patients who might need a CCK prosthesis.