Background. The development of computer-assisted navigation technologies in total knee arthroplasty (TKA) has significantly expanded the possibilities for personalized surgical treatment. Modern navigation systems not only enable highly accurate assessment of knee deformities but also facilitate the implementation of various alignment strategies tailored to individual patient anatomy. In this context, the study of different patterns of knee deformity, including phenotypes classified according to the Coronal Plane Alignment of the Knee (CPAK) system, has become increasingly relevant. The aim of the study — to determine the relationship between coronal knee deformity and flexion contracture, to evaluate deformity correction parameters, and to perform CPAK-based phenotyping using intraoperative computer navigation data in total knee arthroplasty. Methods. A retrospective single-center analysis of 941 TKAs performed with Stryker Navigation between 2008 and 2024 was conducted. Recorded parameters included coronal alignment, maximal extension and flexion, bone resection values, and calculated intraoperative angles (MPTA, LDFA, aHKA, JLO). Using a simplified model, an approximate assessment of the contribution of bone resections and soft tissue release to the correction of deformity in the frontal plane was carried out. The projection of the initial knee joint deformity data onto the CPAK classification was performed. Results. Varus deformity predominated (91%), consistent with global data. A low positive correlation was found between varus severity and flexion contracture (r = 0.25; p 0.001). Postoperative mechanical alignment was achieved within ±2° in most cases, confirming the high precision of navigation. The extent of medial soft-tissue release increased proportionally with deformity severity, reaching the highest values in knees with varus 15°. CPAK phenotype distribution in this cohort differed notably from published radiographic datasets, likely due to the use of intraoperative navigation-derived measurements. Conclusions. Navigation data allow quantitative evaluation of the relationship between deformity, bone resection, and soft-tissue balancing. Computer navigation remains a precise and reproducible tool for deformity correction, while its integration with robotic systems enhances the potential for personalized alignment strategies in total knee arthroplasty.